(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "A summary description of the sources and methods used in estimating county personal income, 1969-74"

^f\.2.-~z.^ ^.jiMus-m, 



A SUMMARY DESCRIPTION OF THE 
SOURCES AND METHODS USED 

ESTIMATING COUNTY PERSONAL INCOME 

1969-74 



^TOr CQ 



c 



..<*x. 



X 






/ 




U.S. DEPARTMENT OF COMMERCE 

Bureau of Economic Analysis 
July 1976 



ATTENTION INCOME DATA USERS! 

Here, for the first time, is a comprehensive and detailed source of personal income data for 
3,618 areas. Prepared by the Bureau of Economic Analysis and published through the 
National Technical Information Service, a 4-volume set titled 



LOCAL AREA 
PERSONAL INCOME, 1969-74 

shows total and per capita personal income by place of residence, personal income by type 
of payment, and labor and proprietors' income by major industry and place of work. 

SPECIAL FEATURE: DETAIL AVAILABLE FOR ALL COUNTIES 



Volume I includes methodology and classification of SMSA's and BEA economic areas. 
Shows data for all States, BEA economic areas, and SMSA's. Accession No. PB 254055. Price 
$10 paper, $2.25 microfiche. 

A summary methodology is included in each of the following volumes. 

Volume II shows data for counties and SMSA's in the following States: Connecticut, Del- 
aware, District of Columbia, Illinois, Indiana, Iowa, Maine, Maryland, Massachusetts, 
Michigan, Minnesota, Missouri, New Hampshire, New Jersey, New York, Ohio, Pennsylvania, 
Rhode Island, Vermont, and Wisconsin. Accession No. PB 254056. Price $13.75 paper, $2.25 
microfiche. 

Volume III covers Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, 
North Carolina, South Carolina, Tennessee, Virginia, and West Virginia. Accession No. PB 
254057. Price $16.25 paper, $2.25 microfiche. 

Volume IV covers Alaska, Arizona, California, Colorado, Hawaii, Idaho, Kansas, Montana, 
Nebraska, Nevada, New Mexico, North Dakota, Oklahoma, South Dakota, Texas, Utah, Wash- 
ington, and Wyoming. Accession No. PB 254058. Price $16.25 paper, $2.25 microfiche. 



The volumes can be ordered from the National Technical Informa- 
tion Service, 5285 Port Royal Road, Springfield, Virginia 22161. Be 
sure to mention the accession numbers of the volumes desired. 



A GENERAL VIEW OF THE ESTIMATES 

PERSONAL INCOME DEFINED 

Personal income is the current income of residents of an area from all sources. 
It is measured after deduction of personal contributions to Social Security, 
government retirement, and other social insurance programs but before deduction 
of income and other personal taxes. It includes income received from business, 
Federal, State, and local governments, households, institutions, and foreign 
governments. It consists of wages and salaries (in cash and in kind, including 
tips and bonuses as well as contractual compensation), various types of supple- 
mentary earnings termed "other labor income" (the largest item being employer 
contributions to private pension, health, and welfare funds), the net incomes 
of owners of unincorporated businesses (farm and nonfarm, with the latter in- 
cluding the incomes of independent professionals), net rental income, royalties, 
dividends, interest, and government and business transfer payments (consisting, 
in general, of disbursements to persons for which no services are rendered cur- 
rently, such as unemployment benefits, Social Security payments, Medicare bene- 
fits, retirement pay of governmental programs, and welfare and relief payments). 

For the measurement of personal income on a regional basis, BEA assigns the in- 
come flows to the State, county, or SMSA in which the individual resides. How- 
ever, BEA also presents labor and entrepreneurial income in industrial detail 
by place of work since the bulk of labor 1/ and proprietors' income is reported 
by industry at the point of disbursement {establishment location) 2/ . This 
income is then adjusted to a place-of-residence basis at an all-industry level. 

A more precise residence adjustment may be achieved by computing adjustment 
factors for each major industry group, thus reflecting industrial differentials 
in commuting flows. The information needed to effect a detailed industry-by- 
industry adjustment, however, is not available. 

Personal contributions for social insurance, which are included in personal 
income as explicit deductions, are also subject to residence adjustment. Since 
employee contributions to most social insurance programs are withheld at the 
source of payroll disbursement, they are estimated, like wages and salaries, 
on a place-of-work basis. Accordingly, most personal contributions must be 
adjusted to reflect place of residence when incorporated into the personal 
income estimates. Several contribution items obtained by place of residence, 
the most important of which are contributions of the self-employed and Govern- 
ment Life Insurance contributions, do not require adjustment. 

Regional measures of labor and proprietors' income are important on both 

a place-of-work and place-of-residence basis. The estimates based on place of 

work (where earned), which include detail by industry, are useful in the analy- 



]_/ Labor income is the sum of wages and salaries plus other labor income. 

2/ In the contract construction industry, point of disbursement may or may not 
be the actual work site. Therefore, the wage and salary estimates for the 
construction industry do not necessarily reflect the county of work. This 
is the only industry where this distinction is of importance. 



sis of the industrial structure of a given area. The estimates based on place 
of residence (where received) are useful for the analysis of consumer markets 
and purchasing power. 

Estimates for the other components of personal income—transfer payments, 
dividends, personal interest income, and rental income of persons--are made on 
a where-received basis only. In the case of transfer payments, there is no eco- 
nomic relevance to the where-earned concept. Dividends, interest, and rent are 
not estimated on a where-earned basis because of the lack of data suitable for 
assigning such income to the areas in which it is generated. Moreover, there 
is also a question as to the economic relevance of measuring the investment 
income components of personal income in this context since they do not repre- 
sent a complete measure of the contribution of capital to production (dividends 
being the only part of corporate profits included). 

ORIGIN AND USE OF COUNTY INCOME ESTIMATES 

BEA's work on regional income estimation began in the midthirties with the con- 
struction of a series of "State income payments" to individuals. State income 
payments were the sums of wages and salaries, proprietors' income, dividends, 
personal interest income, rental income (the last three components, in aggre- 
gate, formerly referred to as "property" income), and "other" income, and were 
produced as part of a broad effort to explain the processes and structure of 
the nation's economy in the thirties. The estimation of State income payments 
was the precursor of the extensive work carried out during the forties and 
early fifties on the more comprehensive annual State-by-State breakdowns of 
total personal income, and of work completed in the sixties on the development 
of quarterly estimates of State personal income. 

Currently, a condensed set of State personal income estimates is prepared 
quarterly. Additional detail on State personal income is prepared annually. 
The quarterly estimates cover the post-World War II period, and the annual 
record extends back to 1929. 

During the latter half of the fifties, BEA inaugurated its sub-State estimates 
and projections of personal income and related measures. During this period, 
the local area measurement effort was limited to the development of estimates 
of personal income and employment for a relatively small number of counties in 
the mideastern and plains States. In the rnidsixties, BEA constructed an his- 
torical series of personal income estimates covering selected years 1929 to 1962 
for the nation's SMSA's and non-SMSA counties. 

BEA is currently producing annual personal and money income estimates for each 
State, SMSA, and county as well as for the parishes of Louisiana, the independ- 
ent cities of Virginia, and the census divisions of Alaska. County personal 
income data are available for the years 1929, 1940, 1950, 1959, 1962, and for 
1965-74. Estimates of per capita personal income by State, SMSA, and county 
have been prepared for the same years as the aggregate personal income measures. 

In this publication BEA is introducing, for the first time in published form, 
a comparable set of State, economic area, metropolitan area, and county esti- 
mates of personal income by type of payment and of labor and proprietors' income 
by major industry group. Similar tables for the earlier years (1929, 1940, 



1950, 1959, 1962, and 1965-68) are available upon request. Samples of other 
local area income tables available upon request are shown as Exhibits A and B 
in the Appendix. 

Regional measures of personal income serve a wide range of uses. Federal, 
State, and local governments use them to analyze economic conditions in various 
areas, to serve as a basis for allocating funds, to monitor the effectiveness 
of government programs, to estimate in advance of adoption the regional effects 
of alternative development programs, to measure the capacity of local areas to 
provide tax revenues, and to gauge differences in welfare. The National School 
Lunch Act and the General Revenue Sharing Act are two examples of programs that 
use BEA's estimates of personal income to allocate funds. Business uses BEA's 
estimates as a direct measure of consumer markets by geographic area and as an 
indirect measure of regional industrial markets. Universities and research 
organizations use these same estimates to identify and measure the factors 
responsible for area differences in levels of income and in rates of economic 
growth. 



DERIVATION OF THE ESTIMATES 

The local area income estimates prepared by the Regional Economic Measurement 
Division are tied directly to BEA's official estimates of personal income by 
States. 3/ That is, the State total for each income component, as taken from 
the official State income series before adjustment for residence, is allocated 
to the counties of the State in accordance with each county's proportionate 
share of the same or some related series that is available on a county basis. 
For some income items, more information is available for SMSA's than for indi- 
vidual counties. In these instances, State totals are first allocated to the 
individual SMSA's of each State on the basis of a particular allocator and the 
remainder assigned to all non-SMSA counties by the most appropriate allocator 
available. Prior to the release of the 1973 estimates, it was not possible to 
provide county breakdowns of multicounty SMSA's due to a lack of sufficient 
information. However, the recent availability of Census commuting flows data 
and IRS county tabulations of selected income items has enabled BEA to make 
estimates of personal income for all counties. 

Because of the availability of State control totals, a different allocator may 
be used in each State without impairing the interstate comparability of the 
estimates. This application of individual State totals makes it possible to 
utilize the best county estimating series in each State. In addition, the 
insertion of comparatively accurate benchmark "controls" at the State level 
insures increased accuracy of sub-State estimates. 

The allocation procedure also makes for greater accuracy in the county estimates 
because most components of personal income can be estimated more reliably for 



3/ A supplement to the Survey of Current Business entitled Personal Income 
by States Since 1929 contains definitions, sources of data, and methods 
of estimation used to construct the State personal income series. Although 
out of print, copies of this volume are available at the libraries and 
Bureaus of Business Research of most colleges and universities throughout 
the country. 



States than for smaller geographic areas. Allocation permits the utilization 
of numerous related series of data which, while approximating, may not precisely 
"match" the basic series to be allocated. 

More than 300 series of separate estimates go into the derivation of the 
30 line items shown in the published personal income tables for the SMSA's and 
counties. This detailed approach provides the information needed to audit both 
the sources and methods of estimation, and assures that the final estimates 
will reflect appropriate interindustry and intercomponent differentials at both 
the State and county levels. While the publication of more detail would enhance 
the usefulness of the estimates, Federal and State regulations established to 
safeguard the privacy of individuals or individual establishments prohibit the 
release of more detailed statistics. 

The bulk of the source materials used to prepare the estimates is culled from 
the administrative records of Federal and State government programs, with the 
remainder of the data coming from the various censuses and from nongovernment 
sources. Several of the more important sources of administrative record infor- 
mation include data generated as the by-product of the State unemployment pro- 
grams of the Employment and Training Administration, the insurance programs of 
the Social Security Administration, and the Federal tax program of the Treasury 
Department. Two of the more important censuses utilized are the Censuses of 
Agriculture and Population. The data obtained from these sources yield more 
than 90 percent of the data needed for the preparation of State and county 
income estimates. 

BEA supplements these basic statistics, which may be presumed to be "reliable," 
with data of lesser quality, scope, and relevance. In order to adjust for the 
gaps and deficiencies in the poorer quality data, indirect procedures and arbi- 
trary assumptions are sometimes required. The use of such indirect procedures 
varies considerably over the 45-year estimating span. Moreover, the impact of 
these procedures on the individual State and county estimates varies in accord- 
ance with regional differences in industrial structure. It is not possible, 
therefore, to provide measures of the error introduced into the income esti- 
mates by the indirect procedures. 

Because the estimates are based on several million administrative records, the 
data are understandably subject to a wide range of errors as they are encoded 
and processed by the agencies administering their respective programs. Since 
it is not possible to verify each individual record or accept the various admin- 
istrative record files at their face value, BEA has developed several computer- 
ized edit routines to locate major errors in the source materials. The routines 
test the change over time in the industrial composition of employment and wages 
in each county as well as the change over time in the county distribution of 
employment and wages in each industry. Without the use of these or similar com- 
puterized techniques, the reliability of the many components of BEA's county 
income series would be greatly reduced. 

Although the use of computerized techniques improves the quality of the personal 
income figures, it is not desirable at the present time to completely automate 
the county income estimation procedures. Since much of the reported statistical 
information is not directly or wholly suited for income measurement, close 
attention must be paid to each input source, not only to identify and correct 



errors but to adjust for differences in definition, to close data gaps, and to 
obtain statistical comparability among geographic units and over time. 

The major alternative to BEA's present approach to income measurement would be 
to collect the necessary information in surveys of income recipients. The sur- 
vey approach would provide data directly suited for the measurement of personal 
income, eliminating the necessity of adjusting for definitional and conceptual 
differences among the various inputs. Unfortunately, the cost associated with 
this alternative would be prohibitive because a wery large sample would be 
necessary to permit reliable local area estimates. On the other hand, the use 
of administrative records is both reliable and economical because the data are 
usually subject to internal review by the agency administering the program, and 
it costs much less to use data collected by other agencies for other purposes 
than to conduct regional surveys. 



Table 1. - Relative Importance of Wage and Salary Disbursements, 
by Component, to Total Personal Income, United States, 1969 and 1974* 



Dollars (000,000) 



Percent of TPI 



1969 

Total personal income 751,425 

Wage and salary disbursements 510,180 

Farms 3,088 

Other industries ]_/ 1,006 

Mining 5,414 

Construction 31 ,041 

Manufacturing 157,616 

Durables (selected) 

Lumber and furniture 6,668 

Primary metals 12,237 

Fabricated metals and ordnance 14,462 

Machinery except electrical 18,130 

Electrical equipment and supplies 16,216 

Motor vehicles and equipment 8,993 

Transportation excluding motor vehicles: 10,885 

Nondurables (selected) 

Food and kindred products 12,725 

Textile mill products 5,647 

Apparel and other fabric products 6,903 

Paper and al 1 ied products 5 ,770 

Printing and publishing 8,660 

Chemicals and allied products 9,736 

Petroleum refining 1,937 

All other manufacturing, n.e.c 18,647 

Transportation, communications and public utilities 37,576 

Railroads 5,936 

Trucking and warehousing 8,744 

Other transportation 8,455 

Communications and public utilities 14,441 

Wholesale and retail trade 83,901 

Wholesale trade 31 ,651 

Retai 1 trade 52 ,250 

Finance, insurance and real estate 25,665 

Services 65,288 

Hotels and other lodging places 3,315 

Business and repair services 13,340 

Amusement and recreation 3,617 

Personal services and private households 9,513 

Professional services , 35 ; 503 

Government 99 , 585 

Federal civilian 25,271 

Federal military 15,380 

State and local 58,934 



1974 


1969 


1974 


1,151,622 


100.00 


100.00 


757,289 


67.90 


65.76 


4,817 


.41 


.42 


1,872 


.13 


.16 


8,903 


.72 


.77 


46,260 


4.13 


4.02 


211,885 


20.98 


18.40 


9,542 


.89 


.83 


17,680 


1.63 


1.54 


18,489 


1.92 


1.61 


26,366 


2.41 


2.29 


21,375 


2.16 


1.86 


12,536 


1.20 


1.09 


11,515 


1.45 


1.00 


16,815 


1.69 


1.46 


7,452 


.75 


.65 


8,361 


.92 


.73 


7,902 


.77 


.69 


11,536 


1.15 


1.00 


13,672 


1.30 


1.19 


2,810 


.26 


.24 


25,834 


2.48 


2.24 


56,939 


5.00 


4.96 


7,900 


.79 


.69 


13,891 


1.16 


1.21 


11,724 


1.12 


1.02 


23,424 


1.92 


2.03 


127,330 


11.17 


11.06 


49,400 


4.21 


4.29 


77,930 


6.95 


6.77 


39,791 


3.42 


3.46 


105,145 


8.69 


9.13 


4,686 


.44 


.41 


21,190 


1.78 


1.84 


5,372 


.48 


.47 


10,472 


1.27 


.91 


63,425 


4.72 


5.51 


154,347 


13.25 


13.40 


36,863 


3.36 


3.20 


18,210 


2.05 


1.58 


99,274 


7.84 


8.62 



* Detail may not add to higher level totals because of rounding. 

1/ Includes agricultural services, forestry, fisheries, and rest-of-world. 



SOURCES AND METHODS OF ESTIMATION ' 

WAGE AND SALARY DISBURSEMENTS 

The definition of wage and salary disbursements is in accord with common usage. 
It includes executives' compensation, commissions, tips and bonuses, and the 
value of payments in kind which represent income to the recipient. They are 
measured before deductions for Social Security contributions, union dues, or 
other purposes. All disbursements in the current period are covered, includ- 
ing any payments retroactive to past periods; that is, retroactive wages are 
counted when paid rather than when earned. 4/ 

The contributions made by employees under the various social insurance programs, 
although counted in wage and salary disbursements, are not part of the personal 
income total. They are excluded by means of the deduction made for "personal 
contributions for social insurance." 

Estimates of wage and salary disbursements (which account for about 65 percent 
of all personal income) are, with perhaps the exception of transfer payments, 
more complete and reliable than those for any other type of income, and, because 
of their sizable weight in the total income flows, impart a large measure of re- 
liability to the estimates of aggregate income. For the private sector of each 
subarea's economy, the payroll component covers employees not only of all non- 
farm business establishments, but also of farms, private households, hospitals, 
and private educational, social service, and nonprofit institutions. Also, all 
government employees are covered by the measure, including those of the State 
governments, local governments, and Federal government (both civilian and 
military) . 

"Covered" Wages and Salaries 

The bulk of the county wage and employment data used to estimate payrolls has 
been obtained from the administrative records of the Employment and Training 
Administration's unemployment insurance program (ES-202). Firms covered by the 
State unemployment insurance regulations submit regular quarterly reports to 
their respective State Bureaus of Employment Security (BES). These quarterly 
reports include wage disbursements by each firm for the reporting quarter as 
well as the monthly number of full- and part-time workers employed. 

As part of its data acquisition program, BEA obtains from each State BES agency 
two-digit level tapes and/or tabulations of quarterly wages and monthly employ- 
ment which, after editing and correcting, form the basis for more than 80 per- 
cent of total wage and salary disbursements. Until 1972, more than half of the 
States excluded from mandatory coverage those establishments with less than 
four employees (referred to as "sma"n firms"). Beginning with the first quarter 
of 1972, the UI coverage of payrolls and employment was extended to cover most 
firms in all but a few industries. The size-of-firm exclusions were eliminated 
entirely, with firms of all sizes falling under the provisions of the Unemploy- 
ment Act. The sole exception is nonprofit organizations, for which mandatory 
coverage extends only to those with four or more employees. 

4/ While the timing of wages when paid is a clear conceptual feature of per- 
sonal income measurement, the difference between wages earned and wages 
paid has been negligible in most years. 



Adjustments to "Covered" Wages and Salaries 

The UI "industry" data reported by the State Bureaus of Employment Security 
have had to be supplemented and modified in several ways for the purpose of 
deriving a complete measure of covered wages and salaries for the county per- 
sonal income series. Most obvious and important was the necessity of adjusting 
for the payrolls of small firms excluded from coverage by the differing size- 
of-firm provisions of the State laws. 

Employees of firms too small for inclusion in the UI program were nonetheless 
covered under the Old Age and Survivors Insurance (OASI) law, the scope of which 
is not conditioned or affected by the size-of-firm factor. Under the OASI pro- 
gram, however, employers' quarterly reports are required to show only taxable 
payroll, not total and_ taxable, as is the case with employers reporting under the 
UI programs. 

Through statistical analysis of the employers' reports for Old Age and Survivors' 
Insurance, it was possible to derive direct measures from the taxable payroll 
of firms with too few employees to be covered by the UI laws. Such payroll 
data were provided BEA, by the Bureau of the Census in cooperation with the 
Bureau of Old Age and Survivors' Insurance, on a two-digit industry basis for 
each State and county covering the first quarter of 1965. 

For that year, the county distributions indicated by the reported quarterly 
figures were used to allocate the State estimates of calendar year payrolls 
(taxable and nontaxable combined) of small firms in each industry. The 1965 
distribution was extrapolated by "covered" payrolls from 1966 to 1971 to obtain 
comparable distributions on an individual industry basis. 

In making these extensions of the reported quarterly data to the later years, it 
was necessary to adjust for any changes in the State laws with respect to the 
size-of-firm coverage provision. 

Both the UI tabulations and the OASI data regularly show minor amounts of pay- 
roll which have not been assigned to any industry. Since the individual clas- 
sification scheme followed in the national, State, and county income estimates 
permits no "unclassified" category, it is necessary to adopt some convention, 
necessarily arbitrary, to allocate such unclassified payroll among the industry 
groups. The procedure for doing this in the State series, so as to achieve con- 
formity with the independent national estimates, is rather complex and need not 
be detailed here. It should be noted, however, that this particular adjustment 
of the reported UI industry data is usually quite small --approximately $800 mil- 
lion, or about one-tenth of 1 percent 5/ , on a national basis in 1974--and that 
it cannot introduce error into the State payroll totals since only the appor- 
tionment by industry of amounts reported for particular States and counties is 
involved. 

Portions of several "covered" industry payrolls still remain outside the scope 
of the State unemployment insurance laws. These excluded elements consist of 
wages and salaries of the Federal Reserve Board, national banks, State banks that 
are members of the Federal Reserve System in New Jersey (prior to 1972), elec- 
tric railways, carrier affiliates in the transportation industry, insurance 



5/ The size of the "unclassified payrolls" has increased somewhat since 1972 
because of the extended coverage to small firms. 



solicitors working on commission, and employees' tips. In some instances, pay- 
rolls of these industrial segments can be estimated quite readily by counties. 
In others, the task is difficult and the results less satisfactory. 

"Noncovered" Wages and Salaries 

At one time or another, industries not covered or partially covered by UI data 
included farms, Federal, State and local governments, railroads, private house- 
holds, hospitals, nonprofit membership organizations, "museums, art galleries, 
etc.," private educational services, forestry, fisheries, and "rest of the 
world." 

Hospitals, educational services, nonprofit membership organizations, museums, 
and forestry and fisheries were covered more extensively by the UI regulations 
in 1972. This extension of coverage represents an improvement in the estimating 
data base for these industries, but additional estimation is still required be- 
cause the new coverage of nonprofit membership organizations does not extend to 
religious organizations or to other nonprofit organizations with less than four 
employees and nonprofit schools other than institutions of higher education. 
The formulation of estimates for each of these industries is covered in the 
subsequent sections. 

Government Wage and Salary Disbursements 

In 1974, government wages and salaries accounted for 20 percent of total pay- 
rolls and 13 percent of personal income. Thus, the adequacy of the government 
payroll component has considerable bearing on the quality of the State personal 
income estimates. 

Payments to civilian employees comprise almost 90 percent of total government 
payrolls in the current period. The statistical basis for estimating civilian 
government wages has generally been good despite variations in the relative 
accuracy of the Federal, State, and local segments. 

Apart from dependency allotments, for which some direct State and county data 
have been available, military payroll disbursements have been estimated by al- 
location of State totals on the basis of the number of personnel stationed in 
each county. While the lack of payroll data by county is a significant limi- 
tation, the basis of allocation in this case is reasonably satisfactory. 

Government payroll information available from both the Federal civilian and 
Federal military establishments, as well as from the units of State and local 
governments, still lags behind the growing flows of local area wage data avail- 
able from administrative record files in the private sector. The reliability 
of the county estimates for government wages may therefore vary in quality from 
State to State. In most cases, it has been possible to compensate for the de- 
ficiencies in the raw materials by using secondary data sources, and BEA has 
stepped up its efforts to obtain county wage data directly from the respective 
State governments in an effort to improve the quality of the county wage esti- 
mates for State and local government units. 



10 

Federal civilian wages and salaries have been covered by the UI program since 
1967. However, Federal Government units in many States report on an agency, 
rather than an establishment, basis. This reporting of Federal civilian wages 
by agency restricts the usefulness of the UI data as a source for estimating 
Federal civilian wages at the county level. At the present time, UI data are 
used directly in preparing the estimates of Federal civilian wages in only 
15 States; for the other States, county allocations are based on Civil Service 
Commission (CSC) employment data. To offset this limitation, the county allo- 
cation is made in considerable detail. The county employment totals of the 
major departments, as reported by the CSC, are weighted by the appropriate UI 
average wages to derive a more reliable county allocator. Although employment 
represents a secondary source, its use for Federal civilian wages was acceptable 
since regional differentials in average wages of Federal agencies outside the 
Washington, D.C., area tend to be minimal. 

State and local government payrolls represent the fastest growing component in 
the government sector, having increased almost tenfold in the last 25 years. 

Benchmark estimates of local government wages and salaries were prepared for 
each county from the local government payroll data reported in the 1967 and 1972 
Censuses of Government. Estimates for the intervening years were the products of 
straight-line interpolation. The 1973 and 1974 estimates were made in two parts. 
For the 372 largest counties, the local government payrolls were obtained from 
the annual Bureau of the Census publication, "Local Government Employment in 
Selected Metropolitan Areas and Large Counties." These 372 counties accounted 
for 71 percent of all local government wages and salaries. The estimates for the 
remaining small counties were made by extrapolating the 1972 benchmark estimates 
by population and using the extrapolated series to distribute the residual State 
control totals (derived by summing the large county payrolls for each State and 
subtracting the aggregates from the State control totals). 

State government payrolls were constructed using the 1967 Census of Governments 
county tabulations of full-time employment of State government units, since the 
1972 Census of Governments did not include the information necessary to prepare 
a new benchmark. 

Military disbursements - With one exception, wage and salary disbursements for 
all industrial components represent gross earnings of employees without deduc- 
tions of any kind. The exception to this general rule is the military payroll, 
which differs significantly in concept from a measure of the gross earnings of 
military personnel stationed in each State or county. Military disbursements 
by local area are derived as the sum of two separate flows: (1) the gross earn- 
ings of military personnel stationed in each State or county less the amounts 
withheld by the government and sent to their dependents or other individuals 
in the form of voluntary allotments of pay or benefits under the government's 
family allowance or dependency assistance programs: and (2) allowances and 
allotments received by individuals residing in the State or county. The second 
flow represents amounts withheld from the pay of military personnel wherever 
stationed. A noteworthy aspect of this item is that it represents an element-- 
the only one--of wage and salary disbursements received by individuals not in 
an employee status. 

In brief, the military payroll component of personal income represents, for 
each year, that part of the State total military gross pay which is disbursed 
to residents of the various counties. 



11 

Little direct data on military payroll disbursements have been available and, 
as a result, county estimates oiF military payrolls had to be constructed from 
military strength data obtained from the decennial Censuses of Population and 
from the U.S. Department of Defense. Military payrolls were allocated in two 
parts--pay to military personnel and allowances and allotments. State totals 
of cash pay and pay in kind (clothing and food) received directly by military 
personnel were distributed among counties in proportion to military strength. 

State totals of allowances and allotments were allocated to counties by 
civilian and military population with the latter given twice the weight of the 
former. Civilian population was included to take account of the substantial 
volume of pay allotments that military personnel remit to their dependents. 
The two-to-one weighting system was derived from tabulations of actual disburse- 
ments on a State basis. 

Private Wage and Salary Disbursements 

Farm wages at the county level were measured by allocating the State totals of 
farm wages, as estimated by the U.S. Department of Agriculture and adjusted to 
BEA definitions, in proportion to the county distribution of cash farm wages 
reported in the 1969 Census of Agriculture. Farm pay in kind was estimated by 
allocating State totals to the counties on the basis of the number of workers 
on a farm 150 days or more in a year, also reported in the Census of Agriculture, 

Forestry, fisheries, private education, and nonprofit membership organizations - 
With the extension of UI coverage in 1972 and the improvement of OASI coverage 
under the elective coverage provisions, the methods for estimating wages in 
these industries have been designed to take advantage of whichever program 
yields the best coverage for a given industry in a given State. 

The 1972-74 county estimates of wages in forestry (accounting for an insignifi- 
cant fraction of all wages in the U.S.) were based on the UI wage distributions. 
Estimates for 1969-71 were made in two parts. For those States with the UI laws 
providing mandatory coverage of firms with one or more employees, estimates were 
based on the UI county distributions. Estimates for all other States were based 
on the county distributions of first quarter wages reported by County Business 
Patterns (CBP). 6/ 

The wage estimates for the fishing industry (also accounting for a very small 
fraction of total wages) were, for the most part, based on UI data. UI wages, 
adjusted for small firms where necessary, served as allocators for 22 States, 
including all the major fishing States except Oregon, Rhode Island, Mississippi, 
and Maryland. Estimates for these four States were based on modified UI data. 
For the few remaining States having a significant fishing industry, CJ3P data 
were used to distribute the State totals. 

Although UI data were used in estimating wages and salaries in private education 
at the State level, it was not possible to use them for the county estimates. 
Since nonprofit schools other than "institutions of higher education" are explic- 
itly excluded from UI coverage, the elementary and secondary schools component 
of private education should be separated out from the remainder of the industry 
in order to have a comparable data base for allocation purposes. This is pos- 
sible at the State level because UI data are available in greater detail (by 



6/ Based on OASI administrative records, 



12 

three-digit industry). At the county level, however, BEA receives wage data 
for two-digit industry groups only. The county estimates were therefore based 
on the CBP first quarter wage distributions. Although nonprofit schools other 
than "institutions of higher education" are excluded from mandatory coverage 
under OASI regulations as well, the degree of coverage is higher because the 
provisions of elective coverage encourage a high rate of participation. 

Nonprofit membership organizations are excluded from the "one or more" manda- 
tory coverage provisions of the 1972 UI extended coverage law. Moreover, em- 
ployees of religious organizations are explicitly excluded from coverage as well, 
Therefore, the UI data were considered unsuitable as allocators for the county 
estimates. CBP coverage, on the other hand, is somewhat better than the UI 
since employees of religious organizations are not specifically excluded and 
the elective coverage provisions of OASI have encouraged a high rate of 
participation. However, under OASI guidelines, the clergy are consiuored to 
be self-employed and, therefore, they are excluded from the CBP data. Although 
this clearly reduces their relevancy as an allocator, in the absence of better 
data, the CBP first quarter wage distributions were used as the basis for the 
estimates of wages and salaries in nonprofit membership organizations. 

Railroads - County estimates of wages in the railroad industry were based on 
the biennial employment series for Class 1 railroads (line-haul and switching 
and terminal companies) prepared by the Association of American Railroads (AAR). 
These AAR data are for selected SMSA counties and account for approximately 
70 percent of total railroad employment. For the remaining counties, AAR's 
residual -counties employment estimate in each State was disaggregated in pro- 
portion to the railroad employment reported by counties in the 1970 Census of 
Population. Estimates for the intervening years were derived by averaging the 
biennial benchmark data. The resulting employment series was used to allocate 
the State totals of wages and salaries in the railroad industry. The 1973 
distribution was also used to allocate the 1974 State totals. 

Private households - Due to a complete lack of any other relevant local area 
statistics, wages and salaries are computed on the basis of Census income and 
employment data. A 1969 benchmark distribution for allocating the State totals 
of wages and salaries received by private household workers was computed as the 
product of the number of private household workers in each county and an aver- 
age earnings estimate. The average earnings of persons employed by private 
households was derived from Census tabulations of earnings of persons by income- 
size class and by industry. Such averages could be prepared only for the State 
and for each SMSA with a population of 250,000 or more. 7/ An estimate of the 
combined average earnings for all areas outside the reported SMSA's was computed 
from the residual resulting from the subtraction of the SMSA data from the State 
totals. This residual average was used for each county lying outside the re- 
ported SMSA's. 

The number of private household workers (other than unpaid family workers) in 
each State and SMSA with a population of 250,000 or more was tabulated directly 
from the 1970 Census. The SMSA employment numbers were disaggregated into 
estimates for the constituent counties on the basis of a Census county distri- 
bution of employment reflecting the total work force (the difference in concept 
is not significant in the case of private household workers). The residual 
employment number for the counties outside the reported SMSA's was also disag- 



7/ New England SMSA's were excluded for reasons explained in the section 
entitled "Classification of SMSA's." 



13 

gregated into its constituent counties by the Census county employment 
distribution. 

The 1969 distribution was used to allocate the State totals for 1970 and sub- 
sequent years. 

Hospitals - The county distributions of hospital payrolls for the years 1969-71 
were based on data presented in the Journal of the American Hospital Associatio n's 
"Hospital Guide Issue" (called "Hospitals" in recent years) . The AHA data on 
payroll expenses were obtained directly from the hospitals in an annual ques- 
tionnaire survey. 

As a result of the 1972 UI coverage extension, Ul-based estimates of wages re- 
ported for this industry are now considered to be complete; therefore, the use 
of the AHA data has been discontinued, and the estimates for the years 1972-74 
reflect the introduction of the UI data. It was not necessary to revise the 
estimates prior to 1972 because of the close similarity between the two sets 
of data. 

Private museums, art galleries, etc. - County estimates of wages and salaries 
were based on first quarter payroll data for this industry as reported in County 
Business Patterns . UI data were not introduced into the estimates for 1972-74 
because substantia; differences between the UI and CBP distributions warrant 
further investigation. 

The rest-of-the-world components of wages and salaries represent wage payments 
to United States residents from international organizations (such as the United 
Nations) and from foreign governments. The State totals were distributed to 
the counties in which the United Nations, embassies, and consulates of foreign 
governments are located. 

Wages in Kind 

The wage and salary estimates for the various States and counties include allow- 
ances for the food, clothing, and lodging paid in kind to employees which repre- 
sent income to them. The concept of valuation is cost to the employer. Market 
value to the employee would be a preferable concept for some purposes, although 
it is more elusive and less subject to quantitative determination. 

This area of wage imputation is rather imprecise and involves a number of diffi- 
cult decisions which can be settled only in a pragmatic fashion. For instance, 
the imputation is confined to food, clothing, and lodging because other types 
of perquisites, such as medical and recreational services, are generally less 
important and cannot be estimated satisfactorily from available data. It is 
frequently difficult, moreover, to determine whether or not a particular type 
of payment in kind clearly represents an addition to cash wages and salaries. 

Payments in kind are a significant element of military wages. They include the 
value of the food and clothing provided enlisted personnel as part of th^ir 
total pay and allowances. The clothing imputation is confined to "standard" 
issues. It does not include clothing and equipment designed for use on special 
duties or under unusual conditions. 



14 

As to other industrial segments of the State and county estimates, wages in 
kind (comprising either food or food plus lodging) are of some significance in 
eating and drinking places, farming, private households (domestic servants), 
water transportation, hotels, private education, nonprofit organizations, and 
hospitals. They are quite minor, however, in other areas of private employment 
and in the government sector apart from the military. 

In terms of national income accounting, the imputation of wages depicts the 
accounts as though the payments in kind had taken the form of cash flows. In 
the simple case of food furnished restaurant workers, the imputation assumes 
that the employer, instead of furnishing his employees with free food, pays 
them corresponding amounts of wages and the employees in turn use these wages 
to buy items previously purchased by the employer. Employees' wages and busi- 
ness sales to consumers (recorded in personal consumption expenditures) are 
raised by equivalent amounts. Omission of the imputation would understate the 
measures of personal income, personal consumption expenditures, and total output, 
It would also understate the real earnings of employees receiving food relative 
to those paid wholly on a cash basis. 

A breakdown between payments in cash and in kind of private nonfarm wages and 
salaries in the "covered" industries is not available for States or counties. 
This is because such a breakdown is not provided in the basic payroll data for 
industries covered by Social Security legislation. The value of income in kind 
is covered in the payroll tabulations relating to both the State unemployment 
insurance systems and Old Age and Survivors Insurance, but is not reported 
separately by employers. Estimates of pay in kind are made for some of the 
noncovered industries at the national and State levels. Those estimates are 
of varying quality. At the county level, except for farms, the State totals 
of cash pay and pay in kind are distributed by the same allocator. 



OTHER LABOR INCOME 

Other labor income consists of supplementary types of labor income paid out or 
accruing in the current period. These include employer contributions to private 
pension, health and welfare, and group insurance plans; compensation for injur- 
ies; pay of military reservists; directors' fees; and several other minor items. 
The pay of members of the military reserve consists of compensation for inactive 
duty training under the various reserve programs. The employer contribution 
component currently accounts for 85 percent of other labor income nationally. 
The reliability of the county estimates for other labor income thus depends 
very largely upon the employer contribution item. Data on the other components 
remain inadequate. 

Since employer contributions to private pension funds are made on behalf of 
employees, they have been distributed to counties on the basis of payrolls. The 
allocations have been carried out in considerable detail because the ratio of 
employer contributions to wages and salaries differs widely by industry. The 
procedure for estimating employer contributions to health and welfare funds fol- 
lowed the same pattern, using employment as the primary county allocator. A simi- 
lar procedure was utilized for estimating compensation for injuries, employer 
contributions to supplementary unemployment benefit funds, and directors' fees. 



15 



Table 2. - Relative Importance of Other Labor Income, by Component, 
to Total Personal Income, United States, 1969 and 1974* 

Dollars (000,000) Percent of TPI 



Total personal income 751,425 

Other labor income 

Compensation for injuries 

Employer contributions to private 

pension plans 

Employer contributions to health and 

wel fare funds 

All other 1/ 



1969 


1974 


1969 


1974 


751,425 


1,151,622 


100.00 


100.00 


28,415 


51,378 


3.78 


4.46 


2,766 


4,651 


.37 


.40 


11,527 


22,132 


1.53 


1.92 


12,118 


21,334 


1.61 


1.85 


2,004 


3,261 


.27 


.28 



* Detail may not add to higher level totals because of rounding. 

]_/ Includes pay of military reservists, directors' fees, compensation of 
prison inmates, marriage fees to Justices of the Peace, jury and witness 
fees, Federal Government contributions to Government Life Insurance, and 
employer contributions to supplementary unemployment benefit funds. 



16 

The remaining minor items of other labor income include compensation of prison 
inmates, marriage fees to Justices of the Peace, jury and witness fees, and 
Federal contributions to group life insurance. Together, these items account 
for less than one-tenth of total other labor income, and have been apportioned 
to the counties in terms of payrolls or civilian population as deemed most 
appropriate. 

PROPRIETORS' INCOME 

Proprietors' income measures the net business earnings of owners of unincorpo- 
rated enterprises. Farmers, independent professional practitioners (such as 
physicians, dentists, and lawyers), entrepreneurs in nonfarm businesses, and 
others in a self-employment status are included in the scope of proprietors' 
income. Two broad segments of this income component may be distinguished with 
respect to source material and methods of estimation--nonfarm and farm pro- 
prietors' income. 

Under business accounting practices generally followed in reporting for tax 
purposes, inventories are charged to cost of sales in terms of original, not 
current, costs. The result of these practices is the inclusion in business 
profit of an element of inventory gain (or loss) due solely to price change 
and therefore akin to capital gain (or loss). This is not suitable for national 
income purposes, which require a measure of business profits accruing from cur- 
rent production. Such a measure is obtained by adding to profits derived from 
tax return tabulations an "inventory valuation adjustment." This adjustment 
represents the difference between the current replacement cost of inventories 
charged to cost of sales and their reported "book" value which, as indicated, 
usually reflects prior period costs. No such valuation adjustment is required 
in the case of farm inventories, since the farm income estimates are not based 
on tax return information and are computed directly so as to exclude inventory 
profit or loss. 

Nonfarm Proprietors' Income 

For the country as a whole, nonfarm proprietors' income is identical with the 
"business and professional" category of table 1.10 appearing in the July issues 
of the Survey of Current Business prior to 1976. The newly revised national 
income and product accounts, published in the January 1976 SCB , Parts I and II, 
introduce an additional modification of nonfarm proprietors' income: the capital 
consumption adjustment. The new format appears in table 1.13, Part II, of the 
January issue. State and local area personal income estimates will reflect this 
and other conceptual changes newly introduced into the national income and prod- 
uct accounts when the Regional Economic Measurement Division's own benchmark 
revisions are released in late 1977 (State estimates) and early 1978 (local area 
estimates) . 

Nonfarm proprietors' income (prior to revision) consists of two items--" incomes 
of unincorporated enterprises" and "inventory valuation adjustment." The former 
items consists wholly of monetary earnings. These accord closely in definition 
with net business profit (gross receipts from business or profession less ex- 



17 



Table 3. - Relative Importance of Proprietors' Income, by Component, 
to Total Personal Income, United States, 1969 and 1974* 



Dollars (000,000) 

1969 1974 

Total personal income 751,425 1,152,622 

Proprietors' income 67,191 87,654 

Farm 16,741 28,154 

Nonfarm 50,450 59,500 

Agricultural services, forestry 

and fisheries 781 885 

Mining 32 125 

Contract construction 5,122 7,217 

Manufacturing 1,794 2,259 

Transportation, communications and 

public utilities 1,338 1,883 

Wholesale and retail trade 13,418 15,566 

Finance, insurance and real estate 4,055 2,912 

Services 23,910 28,653 

Business services 5,790 7,116 

Professional services 18,120 21,537 



Percent of TPI 



1969 


1974 


0.00 


100.00 


8.94 


7.61 


2.23 


2.44 


6.71 


5.17 


.10 


.08 


1/ 


.01 


.68 


.63 


.24 


.20 


.18 


.16 


1.79 


1.35 


.54 


.25 


3.18 


2.49 


.77 


.62 


2.41 


1.87 



* Detail may not add to higher level total because of rounding, 
1/ Less than .005 percent. 



18 

pense of doing business) as reported by individuals and partnerships on their 
Federal income tax returns. 

While separate estimates of "book" profits and inventory valuation adjustment 
are made at the national level, lack of relevant information precludes the devel- 
opment of estimates in similar detail at the State and county levels. 

Nonfarm proprietors' income was based essentially on a 1962 county distribution 
of the all -industry State estimate derived from two sources. About two- thirds 
of the aggregate was allocated by reported Internal Revenue Service (IRS) data 
and the remaining one-third by the product of the number of nonfarm proprietors 
and average wages. This series was disaggregated by industry and extended to 

1973 and intervening years by OASI data on the number of small establishments 
by industry and by county. In each instance, the preliminary county distribu- 
tions were adjusted to equal, when summed, the independently and more accurately 
measured State control totals. The 1973 distributions were used to allocate 

1974 State totals. 

Self -employment income data from OASI administrative records for the period 
1968-73 have been secured from the IRS in cooperation with the Social Security 
Administration. The acquisition of this data file by BEA has made possible 
significant improvements in the county estimates of nonfarm proprietors' income 
at the two-digit industry level. Specifically, the IRS/OASI data file contains 
the self-employment income reported by nonfarm proprietors to IRS when filing 
Schedule SE of Form 1040, the schedule for computing the Social Security self- 
employment tax. The records are identified geographically by a city name, 
State name, and zip code. The sole proprietors' records are coded by IRS with 
the SIC codes. No industrial classification was assigned to the records of 
partners' income. In order to ensure confidentiality, the industrial records 
are summed to county and industry and the data suppressed in those instances 
where there are three or less proprietors or partnerships or where the income 
of one or two establishments might be revealed. 

Tabulations of these data are still being edited and evaluated. The processing 
of the OASI data for the trade and service industries, for the years 1968-70, 
has been completed. Since these two industry groups together account for almost 
75 percent of total nonfarm proprietors' income, the applicable 1968-70 self- 
employment income data were incorporated into the estimates during the recent 
estimating cycle rather than deferring their inclusion until all the data were 
processed. The 1970 benchmark estimates for trade and services at the two-digit 
level were extrapolated forward to 1973 by the OASI small establishment data 
published annually in County Business P atterns and adjusted to the State totals. 
The 1974 estimates were based on the 1973 distributions. 

Fa rm Proprietors' Income 

Estimates of the net income of farm proprietors at both the State and county 
levels are equal to (and derived statistically as) the gross income of all farm 
operators minus production expenses and adjusted to exclude the income of cor- 
porate farms (see Exhibit B). Although unpublished, copies of these tables for 
the various counties are available upon request. 

The concepts underlying the BEA county estimates of farm income are the same 
as those used for the national and State farm income estimates prepared by the 



19 

U.S. Department of Agriculture. The major conceptual difference between the 
two series is that the USDA totals include income of corporate farms while the 
BEA figures exclude income of corporate farms. It should be noted that total 
net farm income estimates measure income arising out of the current year's 
production in the farm sector. 

In order to arrive at this level, income is adjusted by the value of the net 
change during the year in farm inventories of livestock and crops held for sale. 
If farmers sell crops in the current year that were produced in prior years, 
cash receipts from marketing will be overstated by the amount of sales from 
storage, the amount held in inventory will decline, and the net change in value 
of inventories will be negative. Conversely, if farmers store more of current 
output than they sell from storage, cash receipts will be understated and the 
net change in value of inventories will be positive. In either case, the cor- 
rection yields a measure of farm income reflecting the results of current 
production. 

The methods used to generate farm proprietors' income rely heavily on quinquen- 
nial data obtained from the Censuses of Agriculture and on selected intercensal 
data prepared by the Economic Research Service (ERS) at the State level and by 
the Statistical Reporting Service (SRS) at the county level. These data are 
used, where possible, to bridge the intercensal gap. In addition to these 
basic sources, other statistical sources in the U.S. Department of Agriculture, 
such as the Agricultural Stabilization and Conservation Service and the Federal 
Crop Insurance Corporation, are utilized in the preparation of a fairly detailed 
income and expense statement covering all farms in each State and county. 

A substantial amount of farm data is reported to the IRS. However, these data 
are of limited use for local area farm income estimation due to a lack of stan- 
dard accounting procedures on the part of the small independent proprietors. 
The diversity of accounting methods makes it difficult to calculate net income 
on any uniform basis. IRS tabulations have an additional limitation in that 
the tax concepts of income and deductions are not comparable with national in- 
come accounting concepts. Finally, IRS tabulations do not adequately cover the 
incomes of many farmers in the lower income brackets. 

Gross income covers the following separately estimated items: (1) cash receipts 
from farm marketing of crops and livestock, (2) payments to farmers under the 
several government payment programs, (3) the value of food and fuel produced 
and consumed on farms, (4) the gross rental value of farm dwellings, and 
(5) the value of the net change in inventories of crops and livestock. 

Cash receipts from marketings is the most important component of gross income. 
The USDA includes some 150 different commodities and generally has production, 
marketing, and price data available for preparing the estimates on a State basis 
However, with the exception of a few States, annual county estimates of cash 
receipts by component are not available. 

To offset this lack of current county data, estimates of cash receipts from 
marketings have been made by summing the USDA State estimates of cash receipts 
for individual commodities into the groups for which cash marketing data have 
been reported by county in the quinquennial Census of Agriculture. These aggre- 
gates were then allocated by a 1969 Census county distribution based on value 
of sales. In order to take account of the specific influence of weather upon 



20 

crop production, supplemental county estimates of annual production of selected 
field crops, available from the SRS, have been used to extrapolate the related 
Census of Agriculture benchmarks. In addition, SRS county data on production or 
cash receipts for commodities other than field crops, when available, have also 
been used as extrapolators. The county estimates for 1971-73 include for the 
first time county data on cash receipts from marketing crops prepared by the 
State SRS offices in Illinois, Montana, North Carolina, Ohio, Pennsylvania, 
South Dakota, and Texas. County data from three additional States — Colorado, 
Arizona, and Oregon—were introduced into the 1974 estimates. 

We had assumed in the past that the relationships among counties with respect 
to livestock tend to be fairly constant from one census to the next and that 
marketings in the counties will move up or down together. For some of the 
important livestock-producing States, however, annual SRS county estimates of 
livestock on feed or cattle and calves on farms as of January 1 are now used 
to extrapolate cash receipts from the latest Census benchmark distributions. 
Moreover, the county estimates of cash receipts from the marketing of livestock 
prepared by six State SRS offices—Illinois, Montana, North Carolina, Ohio, 
Pennsylvania, and South Dakota—were first incorporated into the estimates for 
1971-73. The 1974 estimates include data from Arizona and Oregon, as well. 
These changes were dictated by intercounty shifts in livestock production. 

All State and national estimates prepared by the USDA, with the exception of 
livestock marketed and purchased and rent paid, are used as control totals in 
making county allocations. However, because of differences in timing between 
the farm income and other components of personal income, it is not always pos- 
sible to incorporate the most current USDA totals. When this situation arises, 
the national and State farm control totals are estimated in-house, using USDA 
and BEA data to ensure a consistent series. Because interfarm, intra-State 
transfers are compensating income and expense items when State accounts are 
aggregated } the USDA excludes these transactions in its estimates of cash re- 
ceipts from marketing livestock and expenses of livestock purchased. However, 
at the county level, such transactions are not necessarily between farmers 
within a county, and an estimate of both income and expenses (equal at the 
State level) is made by BEA and distributed among counties by two different 
allocators— cash receipts for livestock and expenses for livestock purchased. 
Similarly, for net rent, the allocators for rent received are different from 
the allocators for rent paid. 

Estimates of total government payments to farmers are available by county from 
the Agricultural Stabilization and Conservation Service (ASCS) on an annual 
basis. 

There is relatively little direct information available for estimating county 
totals of such nonmoney income items as value of home consumption, gross rental 
value of farm dwellings and inventory change. The county estimates for these 
items are, for the most part, based on related data obtained from the Census 
of Agriculture. 

Estimates of the change in farm inventories are weak. Production data were 
used to allocate inventory change in the same detail that State estimates were 
available. SRS production data were used when and where available. A seri- 
ous drawback in using this technique is that it imposes upon each county the 
change registered at the State level, even though some counties may change in 
the opposite direction. 



21 

Production expenses were estimated by counties for some 45 items using direct 
and indirect county allocators to distribute the USDA-based State totals. 
Expenses of farmers by county for items comprising, on the national level, 
over half of total production expenses were available from the 1969 Census of 
Agriculture. These items are: feed purchased, seed purchased, expenditures for 
petroleum fuel and oil, livestock and poultry purchased, fertilizer and lime pur- 
chased, expenditures for pesticides, and cash outlays for hired farm workers. 
Data on the above items were used directly as reported, adjusting to the USDA 
levels when necessary. (Annual SRS county estimates of cash receipts from live- 
stock and products were used to extrapolate the Census benchmark estimates of 
feed and livestock purchases in Illinois, Montana, North Carolina, Ohio, 
Pennsylvania, Arizona, and Oregon.) 

The remaining production expenses itemized at the State level were allocated 
to counties by using indirect, but related, Census data such as crop produc- 
tion, machinery on farms, acres in farms, value of land and buildings, etc. 

Because personal income refers to only that part of farm net income accruing 
to sole proprietors and partnerships, whereas the methodology used (as described 
above) estimated the net farm income of all farms, a further adjustment was made 
to exclude corporate farms. 

The method introduced in the 1973 estimating cycle (and incorporated into the 
system for prior years) splits the aggregate net income of all farms in a given 
county on the basis of the division of acreage between corporate and noncor- 
porate farms in that county, as reported in the 1969 Census of Agriculture. 
The resulting distributions were then scaled to State levels. Using this pro- 
cedure, the net incomes of corporate and noncorporate farms were both either 
positive or negative, depending upon whether the estimate for all farms was 
a profit or loss. 

DIVIDENDS, INTEREST, AND RENTAL AND ROYALTY INCOME 

Dividends, personal interest income, and rental and royalty income of persons, 
both monetary and imputed, have maintained a relatively stable 12 to 14 percent 
share of the total personal income over the last 25 years. 

The quality of the estimates of dividends and monetary interest has improved 
significantly in rpcent. years with the introduction of comprehensive and direct 
data at the county level made available by the internal Revenue Service (IRS). 
Prior to this, only indirect, incomplete and/or summary data were available for 
most counties. Monetary rents, royalties, and the imputed items (imputed interest 
and imputed rent), which together represent the remaining 40 percent of this type 
of payment, are still based on partial or incomplete data. Because of this, the 
county estimates of rental income as well as estimates for the imputed items 
continue to be of a relatively low order of reliability. 

The monetary income components include dividends, interest, and rent and royal- 
ties received by individuals, fiduciaries, and nonprofit institutions. At the 
State level, it is possible to make separate estimates tor each category of 
recipient. Because of the lack of detailed sub-State information pertaining 
to the income receipts of fiduciaries and nonprofit institutions, county esti- 
mates of dividends, monetary interest and monetary rent contain no breakdown 



22 



Table 4. - Relative Importance of Dividend, Interest, 
and Rental and Royalty Income, by Component, to Total Personal Income, 

United States, 1969 and 1974* 

Dollars (000,000) Percent of TPI 



1969 

Total personal income 751,425 

Dividends, interest, and rents 

and royalties 106,147 

Dividends . .. '. . 24,331 

Interest 59,265 

Monetary 35 ,256 

Imputed 24,009 

Rents and royalties.. 22,551 

Monetary 9,502 

Imputed 13,049 



1974 


1969 


1974 


1,151,622 


100.00 


100.00 


163,000 


14.13 


14.15 


32,700 


3.24 


2.84 


103,800 


7.89 


9.01 


70,197 


4.69 


6.10 


33,603 


3.20 


2.92 


26,500 


3.00 


2.30 


7,595 


1.26 


.66 


18,905 


1.74 


1.64 



Detail may not add to higher level totals because of rounding 



23 

of income flows by category of recipient. The proportion of monetary property 
income received by fiduciaries and nonprofit insitutions is estimated at about 
10 percent. The method of estimation places most of the income of these quasi- 
individual s in the more urbanized or higher-income counties. 

Dividends, Monetary Interest, and Monetary Rents and Royalties 
Dividends and Monetary Interest 

Separate special tabulations by county of dividend and interest income as re- 
ported on individual income tax returns for 1972 have recently been made avail- 
able to BEA by IRS, enabling the establishment of reliable and more current 
benchmark estimates. Similar county tabulations are expected for 1974 and 
biennially thereafter. Five-digit zip-coded tabulations of dividend and inter- 
est income were also provided by the IRS for 1969. The zip-coded data required 
conversion to a county basis, posing some problems where a zip code overlapped 
more than one county. The summation of the five-digit zip code to county levels 
produced a 1969 county distribution that is somewhat less reliable than the 
1972 estimates which were based on actual county tabulations. 

Benchmark estimates of total dividends and total monetary interest for 1969 and 
1972 were produced by disaggregating State control totals for each income flow 
by the corresponding 1969 and 1972 IRS-based county tabulations. The 1973 esti- 
mates were made by disaggregating the State totals by county distributions de- 
rived by extrapolating the 1972 benchmarks to 1973 by per capita personal income 
modified to exclude dividends, interest, rent, and farm proprietors' income. 
The 1974 State control totals were allocated to counties by the 1973 distribu- 
tions because of the lag in the availability of the IRS data. Estimates for 
1970 and 1971 were derived by interpolating between the 1969 and 1972 benchmarks 
by the annual change in the modified per capita personal income for each county. 
The interpolated series for each income flow was used to distribute the relevant 
State control totals. 

Monetary Rents and Royalties 

Since there are no five-digit zip-coded or county tabulations of rental and 
royalty income available from IRS, the 1969-74 estimates of monetary rents and 
royalties were developed, in part, directly from the IRS sample data for the 
top 125 SMSA'r- 8/ ds published biennially in the Statistics of Income (SOI) 
and, in part, on the relationship between monetary rents dnc\ royalties and per 
capita personal income less farm proprietors' and investment income. For the 
benchmark years, thp county distribution of monetary rent was b^pd on the 
assumption that it paralleled that of monetary interest. The Statistics of 
Income SMSA data on rents for 1967, 1969, and 1971 were disaggregated to trie 
county level on the basis of the 1969 benchmark for monetary interest. The 1969 
distribution of monetary rent was subsequently adjusted for sampling variability 
by averaging the 1967, 1969, and 1971 distributions. The derived 1969 estimates, 
adjusted to BEA levels, provided the final estimates for the constituent counties 
of the selected iarge SMSA's. The 1970-73 estimates for these counties were 
produced by extrapolating the 1969 final estimates forward to 1973 by the change 
in per capita personal income modified to exclude farm proprietors' income, divi- 
dends, interest, and rent. 



8/ New England SMSA's were excluded for reasons explained in the section, 
"Classification of SMSA's." 



24 

Estimates of monetary rent and royalty income for the remaining counties for 
1969 and 1972 were prepared by disaggregating the "residual counties" control 
totals (derived as the difference between the State control total and the sum 
of the final estimates for the SMSA counties) by the corresponding 1969 and 

1972 benchmark distributions of monetary interest. Estimates of rental income 
for 1970, 1971, and 1973 were developed by extrapolating, or interpolating 
between, the 1969 and 1972 final distributions by the change in the modified 
per capita personal income. The resulting series were used to allocate the 
"residual counties" control totals for the relevant years. Because of the lack 
of current county information on monetary rents and royalties, the 1974 esti- 
mates were made by allocating the 1974 State control totals on the basis of the 

1973 county distributions. 

The foregoing procedure was followed for all States except California. The 
California Franchise Tax Board in its "Annual Report" publishes county data on 
income and losses from rents and royalties. County distributions of net in- 
come from rents and royalties derived from these data were used as allocators 
of the California State control totals. 

Furthermore, 1972 rental incomes in New York and Pennsylvania were adjusted 
to include losses due to flood damage incurred from Hurricane Agnes. County 
estimates of rental loss in these two States were developed by distributing the 
State totals of rental loss by data on total assistance to counties damaged by 
flood, which were made available by the Small Business Administration. This 
distribution of rental loss was subtracted from a recomputed county distribution 
of monetary rents and royalties (which had been constructed by allocating a 
State control total, in which the rental loss was added back) proportionate to 
the original unadjusted county distribution of monetary rents and royalties. 

Imputed Interest and Imputed Rent 
Imputed Interest 

Imputed interest represents the excess of income received by financial inter- 
mediaries from funds entrusted to them by persons over income disbursed by these 
intermediaries to persons. Part of imputed interest reflects the value of 
financial services rendered persons by financial intermediaries without charge. 
The remainder is the property income withheld by life insurance companies and 
mutual financial intermediaries on the account of persons, such as the addi- 
tion of income to policyholder reserves held by life insurance companies. 

In the absence of any information reflecting the amounts of imputed interest 
accruing to residents of the various counties, State totals of this item for 
1969-74 were allocated by the county estimates of monetary interest. 

Imputed Rent 

Imputed rent measures the gross rental income accruing to nonfarm residents of 
owner-occupied nonfarm dwelling units less the normal expenses incurred in home 
ownership. A similar imputation for farm dwellings is implied in the estimates 
of the net income of farm operators. County estimates of nonfarm imputed net 
rent were prepared by allocating State totals to the county by the estimated 



25 

market value of owner-occupied, single-family nonfarm homes. This estimated 
market value was prepared for 1970 by multiplying the number of owner-occupied, 
single-family nonfarm dwellings in each county by their median value. Both 
numbers of houses and median values were reported in the 1970 Census of Housing. 
The 1970 amounts were used to allocate State totals for 1969 and subsequent 
years. 

As in the case of monetary rental income, special estimating procedures were 
used for New York and Pennsylvania in 1972 to take account of the losses in- 
curred due to the flood damage resulting from Hurricane Agnes. County estimates 
of imputed rental loss were derived by allocating the State totals of imputed 
rental loss by the data on total assistance to counties damaged by flood (as 
reported by the Small Business Administration). This distribution of imputed 
rental loss was subtracted from a county distribution of unadjusted imputed rent 
(i.e., unadjusted for losses from flood damage) to yield the final estimates of 
imputed rent for the counties of New York and Pennsylvania. The county distri- 
bution of unadjusted imputed rent for each State was constructed by allocating 
a State total (the sum of imputed rent and the absolute value of the imputed 
rental loss) by the Census-based county distribution of the market value of 
owner-occupied, single-family nonfarm homes. 

TRANSFER PAYMENTS 

Transfer payments, one of the fastest growing components of personal income, 
are, in general, receipts of persons from government and business (other than 
government interest) for which no services are rendered currently. The estimates 
of total transfer payments represent the summation of over 50 separate series. 
Many of the components are the result of detailed data collection. Others are 
estimated by means of allocators which vary considerably in quality. Nationally, 
data from administrative or fiscal records underlie approximately three-fourths 
of total transfer payments. At the county level, the proportion of total trans- 
fer payments based on the fiscal records of government agencies may vary from 
50 to 80 percent, depending on the importance of the income flow to individuals 
from such programs as Old Age and Survivors Insurance, State unemployment insur- 
ance, Medicare, and various welfare and relief programs. 

Federal Government 
Benefits from Social Insurance Funds 

Old Age, Survivors and Disability Insurance (0ASDI) 

County data on benefit payments disbursed by States under the Federal Old Age 
and Survivors Insurance program were tabulated from information reported by the 
Social Security Administration (SSA). 

The total cash benefits paid during the year include monthly benefits paid 
to retired workers, dependents and survivors, and special payments to persons 
72 years of age and over; lump sum death payments; and disability payments to 
workers and their dependents. 



25 



Table 5. - Relative Importance of Transfer Payments, by Component, 
to Total Personal Income, United States, 1969 and 1974* 



Dollars (000,000) 



Percent of TPI 



Total personal income 

Transfer payments 

Federal Government 

Benefits from social insurance 

0ASDI benefits 

State unemployment insurance benefits 

Unemployment compensation for Federal employees 

Railroad benefits 

Government life insurance benefits 

Federal civilian pensions 

Medicare benefits 

Veterans' benefits 

Veterans' pensions and compensation 

Unemployment insurance benefits for veterans 

Veterans' readjustment benefits 

Other V 

Black Lung benefits (1970) 

Educational assistance 2/ 

Public assistance 

Food stamps 

Supplemental Security Income (1974) 

Basic benefits 

"Hold-harmless" 

Other 3/ 

Al 1 other Federal Government transfers 4/ 

State and local government 

Benefits from social insurance 

State and local government retirement benefits , 

Cash sickness benefits 

Public assistance , 

Direct relief , 

Aid to Families with Dependent Children 

General assistance 

Supplemental Security Income: State Supplementation (1974) 
State and local Medicare benefits .. 

All other State and local transfer payments 6/ 

Business 7/ 



1969 


1974 


1969 


1974 


751,425 


1,151,622 


100.00 


100.00 


65,768 


140,092 


8.75 


12.16 


50,338 


114,632 


6.70 


9.95 


40,049 


86,846 


5.33 


7.54 


26,381 


57,626 


3.51 


5.00 


2,102 


6,550 


.28 


.57 


47 


154 


.01 


.01 


1,634 


2,839 


.22 


.25 


744 


817 


.10 


.07 


2,559 


6,355 


.34 


.55 


6,582 


12,505 


.88 


1.09 


8,238 


15,966 


1.10 


1.39 


5,050 


6,984 


.67 


.61 


86 


228 


.01 


.02 


705 


3,381 


.09 


.29 


2,397 


5,373 


.32 


.47 


- 


952 


- 


.08 


320 


767 


.04 


.07 


305 


7,808 


.04 


.68 


237 


3,430 


.03 


.30 


- 


4,242 


- 


.37 


- 


4,136 


- 


.36 


- 


106 


- 


.01 


68 


136 


.01 


. 1 


1,426 


2,293 


.19 


.20 


11,593 


20,252 


1.54 


1.76 


3,771 


8,228 


.50 


.71 


3,415 


7,705 


.45 


.67 


356 


523 


.05 


.04 


6,685 


9,878 


.89 


.86 


6,603 


8,720 


.88 


.76 


5/ 


7.922 


5/ 


.*69 


5/ 


798 


5/ 


.07 


- 


959 


- 


.08 


82 


199 


.01 


.02 


1,137 


2.146 


.15 


.19 


3,837 


5,208 


.51 


.45 



* Detail may not add to higher level total because of rounding. 

1/ Includes military retirement, payments to paraplegics, payments to war orphans, payments to children 
of disabled veterans, payments for automobile and other conveyances for disabled veterans, and educa- 
tional assistance to wives and widows of veterans. 

2/ Includes Federal fellowship payments, Area Redevelopment Act benefits, Manpower Development and Training 
Act benefits, Job Corps benefits, interest subsidy payments on higher education loans, Education Exchange 
(1970), and Basic Education Opportunity Grants (1973). 

3/ Includes Bureau of Indian Affairs' payments and refugee assistance payments. 

4/ Includes Panama Canal Construction Annuity Act payments, Alaska Native Claims Settlement Act benefits 

(1974), Federal payments to nonprofit institutions, Federal auto depreciation payments, and trade adjust- 
ment assistance. 

5/ Prior to 1974, estimated in combination with other types of public assistance to derive a single estimate 
of "Direct relief." (See text, p. 35.) 

6/ Includes veterans' aid, veterans' bonuses, payments to nonprofit institutions, public foster home care 
payments, and State and local government auto depreciation payments. 

7/ Includes consumer bad debts, corporate gifts to nonprofit institutions, cash prizes, unrecovered thefts 
of cash and capital assets, and personal injury payments from business to other than employees. 



27 

Estimates of total OASDI benefits at the county level were based on SSA tabula- 
tions of the amount of monthly benefits paid to those in current payment status 
as of December 31 . 

Special procedures were followed for estimating benefits for Alaska's census 
divisions. An allocator was constructed, based on 1970 Census of Population 
data, by multiplying the number of families receiving Social Security income 
in each census division by the mean amount of Social Security income receiyed. 

State Unemployment Insurance Benefits 

State totals were allocated to counties by benefit data supplied by the State 
Employment Security Offices. While most States report benefits directly by 
county, a few report by local office. In the latter case, local office sta- 
tistics were distributed among the counties that fall within the jurisdiction 
of the district office on the basis of Census of Population data on county 
unemployment. This is a weak procedure, and efforts are being made to obtain 
the more current county unemployment totals being developed by the Employment 
and Training Administration as part of the administrative record requirement of 
the comprehensive Employment Training Act. 

Unemployment Compensation for Federal Employees 

Employment Security Offices in a number of States now supply county or local 
office data on unemployment compensation paid to Federal employees. Where county 
data are available, State totals were allocated directly. Local office data were 
disaggregated using the procedure outlined for "State Unemployment Insurance 
Benefits." For the remaining States, estimates of Federal civilian wages and 
salaries were used as county allocators. 

Railroad Benefits 

Five types of benefits — retirement, survivors, unemployment, cash sickness,, 
and maternity--are paid out under the Railroad Retirement Act and the Railroad 
Unemployment Insurance Act. 

For the period 1969-74, benefit payments made under the provisions of the Rail- 
road Retirement Act were based on a 1971 benchmark estimate constructed from 
1971 zip code data compiled by the Railroad Retirement Board. The zip code data 
were summed to county totals and extrapolated forward to 1973 and back to 1969 
by county estimates of railroad wages and salaries. The 1973 estimates were used 
to distribute the 1974 State totals. Benefits paid out under the provisions of 
the Railroad Unemployment Insurance Act were estimated by allocating State 
totals by county estimates of railroad wages and salaries. 

Government Life Insurance Benefits 

This series is composed of (1) death benefits paid under the National Life 
Insurance Act to survivors of veterans of World War II, the Korean Conflict, and 
the Vietnam Era, (2) death benefits paid from the Government Life Insurance Fund 



28 

to survivors of World War I veterans, and (3) special dividends disbursed to 
veterans holding National Service Life Insurance policies. 

County estimates of Government Life Insurance benefits (including Military and 
Naval Insurance and Servicemen's Indemnity) for the period 1969-74 were based 
on a county distribution of total veteran population as reported periodically 
by the Veterans Administration. (Benchmark data are available for 1969, 1973, 
and 1974. For the 1969-73 period, annual estimates of veteran population were 
derived by interpolating between the benchmarks.) 

Federal Civilian Pensions 

This component includes payments made to, or on behalf of, former employees of 
the Federal Government covered by the Civil Service Retirement and Disability 
Fund or by special contributory and noncontributory retirement systems. 

Federal civilian pensions for the period 1969-74 were estimated in three seg- 
ments—employee annuities, survivor annuities, and all other. Zip code dis- 
tributions of benefit payments for 1971, 1972, and 1973, compiled by the Civil 
Service Commission, were summed to county totals and used as allocators of State 
totals for those years. The 1973 distributions were also used for the 1974 
estimates. For years prior to 1971, county allocators were prepared by extrapo- 
lating the 1971 Civil Service distributions of benefit payments back to earlier 
years by the change in Federal civilian payrolls. 

Medicare Benefits 

Included in this category are the benefits received from both the hospital 
insurance and supplementary medical insurance provisions of Medicare. For all 
States except Alaska, the county distributions of Medicare benefits were based 
on the dollar amount reimbursed by Medicare to persons for medical and hospital 
expenses incurred as reported by the Social Security Administration. Due to 
the lag in the availability of data, the 1973 distributions were used to allocate 
the 1974 State totals. 

The amount of Medicare benefits paid to persons residing in Alaska's census 
divisions for the period 1969-74 were based on estimates of income received 
from Social Security, derived for each division by multiplying the number of 
families receiving Social Security income by the mean amount of Social Security 
income received in 1969 as reported in the 1970 Census of Population. 

Vetera ns' Benefi ts 

Veterans' Pensions and Compensation 

This item consists primarily of compensation to veterans for disability and 
payments to their survivors, including "survivors' indemnity payments" to sur- 
vivors of veterans who were in the Armed Forces on or after June 27, 1950. 
Those eligible for benefits include veterans of all wars since the Spanish- 



29 

American War, as well as those who served in peacetime and incurred service- 
connected disabilities. Veterans with nonservice-connected disabilities who 
are permanently and totally disabled and meet specified income requirements 
are also eligible for benefits. 

An allocating series was constructed from a 1972 zip code distribution of vet- 
erans' pension payments for one month, compiled by the Veterans Administration. 
The zip code data were summed to the county level and extrapolated back to 1969 
by the change in the county distribution of the total veteran population as 
reported by the Veterans Administration (see "Government Life Insurance 
benefits"). This series was used to disaggregate the State totals of veterans' 
pensions and compensation for the years 1969-72. The 1972 distribution was 
used to disaggregate the 1973 and 1974 State totals. 

Military Retirement Benefits 

An allocating series was constructed for this component from the June 1972 and 
June 1974 zip code distributions of military retirement payments compiled by 
the U.S. Department of Defense. The 1972 zip code data were summed to counties 
and extrapolated back to 1969 by the change in the county distributions of 
civilian population. (County estimates of civilian population for 1970 were 
derived by subtracting Armed Forces population from total population as reported 
in the decennial Census of Population. For 1971 and 1972, Bureau of the Census 
postcensal county estimates of population were adjusted to exclude the military, 
using BEA estimates.) State totals of military retirement benefits for 1969-72 
were then disaggregated to constituent counties by this derived allocating 
series. The 1974 zip code data (summed to county totals) were used to allocate 
the 1973 and 1974 State totals of benefits paid. 

Payments to Paraplegics 

This item consists primarily of grants to veterans with service-connected dis- 
abilities requiring specially adapted "wheelchair homes." It applies to all 
veterans since World War II. Due to the lack of more relevant data, county 
estimates have been based on county distributions of total veteran population 
obtained from the Veterans Administration. 

Payments for Autos and Conveyances for Disabled Veterans 

This component covers payment to veterans with specified, service-connected 
physical handicaps toward the purchase price of an automobile (or other convey- 
ance) and necessary adaptive equipment, as well as for the repair, reinstalla- 
tion, or replacement of such equipment. As is the case of payments to para- 
plegics, county estimates have been based on the distribution of total veteran 
population because of the lack of more direct data. 

Veterans' Readjustment Benefits 

The benefits included in this component are subsistence payments for schooling 
and educational allowances made to veterans of the post-Korean Conflict period 
under the Veterans' Readjustment Benefits Act of 1966. 



30 

County estimates of veterans' readjustment benefits were assumed to be in pro- 
portion to the county distributions of veterans of the post-Korean Conflict 
period as reported by the Veterans Administration. 

Unemployment Allowances for Veterans (UCX) 

This item covers unemployment compensation payments to veterans who, in general, 
had 90 days or more of continuous active service and are not receiving certain 
educational assistance or vocational subsistence allowances from the Veterans 
Administration. Although the amount and duration of payments are governed by 
State laws, the benefits are paid from Federal funds. 

The Employment Security Offices in slightly less than half of the States cur- 
rently provide county or local office data on unemployment compensation paid to 
veterans. A varying number of States have been supplying such UCX data since 
1965. Where county benefit data were available, State totals were allocated 
directly. "Local office" data were processed, as described in "State Unemploy- 
ment Insurance Benefits," in order to develop a comprehensive county allocator. 
For the "nonreporting States," the State totals were apportioned among the 
counties in accordance with total veteran population as published by the 
Veterans Administration. 

Other Veterans' Benefits 

This category, comprising less than 10 percent of total veterans' benefit pay- 
ments, includes payments to war orphans, payments to children of disabled vet- 
erans, and educational assistance to wives and widows of veterans. 

Because of the lack of more relevant data, the county estimates of all three 
of these components were made by distributing the State totals in proportion to 
estimates of civilian population for corresponding years. (See "Military 
Retirement Benefits" for the description of the method used for estimating 
civilian population.) 

Other Federal Government Transfers 

While other types of transfer payments are made by the Federal Government in 
addition to social insurance and veterans' benefits, they currently represent, 
in aggregate, less than 5 percent of all Federal benefit payments. A descrip- 
tion of the methods of estimation at the county level of these "other" Federal 
transfer payments are detailed below in order to provide a more complete picture 

Federal Fellowship Payments 

This component includes only the subsistence portion of the fellowship, which 
is paid directly to the individual. The far larger portion is given by the 
government directly to the school and, therefore, is treated in the national 
accounts as either a transfer payment to a nonprofit institution or a govern- 
ment grant-in-aid. Separate estimates are made for Atomic Energy Commission 



31 

(AEC) fellowships, National Science Foundation (NSF) fellowships, payments to 
State Marine School cadets, and "all other" fellowships. 

Since 1970, annual data on the dollar amounts of newly awarded AEC fellowships by 
institution have been available from NSF. The county in which each institution is 
located was identified, and the data were then summed to form a county allocating 
series which was used to distribute the State totals of AEC fellowship subsis- 
tence payments for the period 1970-74. Due to the lack of more pertinent data, 
State totals for 1969 were allocated to counties by civilian population for the 
corresponding years. Civilian population was used for all years for Alaska. 

NSF fellowships are granted to outstanding science students to encourage careers 
in science and engineering. Annual NSF data on the number of students receiving 
such fellowships, by institution, were summed to counties and used to allocate 
the State totals of NSF fellowship subsistence payments for the period 1969-74. 

The Federal Government makes subsistence payments to cadets attending State 
Marine Schools as authorized by the Maritime Academy Act of 1958. Since there 
is only one such school in each of the six States involved, the county estimates 
were made by assigning the State totals to the county in which each school is 
located. 

County estimates of subsistence payments of "all other" fellowships were 
derived by apportioning the State totals in accordance with the county distri- 
bution of the civilian population for the corresponding years. 

Alaska Native Claims Settlement Act Payments 

The Alaska Native Claims Settlement Act, enacted in 1971, provides "for the 
fair and just settlement of all claims by Natives and Native groups in Alaska, 
based on aboriginal land claims." Settlement is made through the granting of 
land rights to groups of Natives and through money payments to resident and non- 
resident Natives. (The claimant, however, must have been a resident of a non- 
urban village comprising at least 25 Natives at the time of the 1970 Census 
enumeration. ) 

Estimates of these payments at the Alaska census division level for 1971-74 
were made by allocating the State control totals by the distribution of the 
number of Natives as reported in the 1970 Census of Population. 

NOTE: The transfer payment component, "Alaska Native Service," which appears 
in the accounts in prior years, and estimated as an insignificant $1 million 
annually, refers to a different Federal program. The Alaska Native Service was 
designed to raise the educational level of the Alaska Natives as well as to 
assist in improving their economic and social conditions. 

Bureau of Indian Affairs (BIA) Payments 

Included in this category are a variety of programs for the benefit of Indians 
on reservations, such as general assistance to needy Indians, adult vocational 
training, and direct aid to college students. County estimates were based on 
1962-63 data on U.S. Indian population on land administered by the Bureau of 
Indian Affairs. 



32 

More current data on the Indian population on reservations are being made 
available by BIA and will be incorporated into the estimates at a later date. 

Alaska totals were distributed among the census divisions by the 1970 Census 
of Population data on Indian population. 

Federal Auto Depreciation Payments 

These are payments for depreciation of privately owned automobiles used in 
Federal Government service. In making the county estimates, State totals for 
all years were allocated to counties by REMD estimates of Federal civilian 
payrolls . 

Food Stamp Act Payments 

This item reflects the net value of the bonus coupons issued to qualifying 
household units. Selected fiscal year county data on bonus coupons issued, as 
reported by the U.S. Department of Agriculture, are available for most States. 
These data were converted to a calendar year basis and then adjusted to State 
totals. For those States for which there are only partial or no county data 
available, allocations were made on the basis of direct relief payments 
reported on a county-by-county basis (see State and Local Government Transfer 
Payments) . 

Refugee Assistance Program 

This component represents the financial assistance for resettlement paid mainly 
to Cuban refugees, as authorized by the Immigration and Refugee Assistance Act 
of 1962. (Similar assistance to Vietnamese refugees, authorized by the Indo- 
china Refugee Assistance Act of 1975, will be reflected in the 1975 estimates.) 

Direct county data have been available from most State Departments of Public 
Welfare. For those States from which no such data were available, civilian 
population was used as the county allocator. 

Black Lung Benefits 

These benefits are the monthly cash payments (beginning in 1970) paid to coal 
miners who are totally disabled by "black lung" disease. Payments are also 
made to dependents as well as specified survivors of miners who died due to 
the disease. Beginning in 1974, regional benefit data for the month of June 
became available from the Social Security Administration. These data were con- 
verted to a county basis for the 15 States in which 95 percent of the benefit 
payments are made. For the remaining States, State totals were allocated to 
counties by the number of persons 60 years of age and over receiving OASDHI 
benefits. 



33 

Basic Education Opportunity Grants 

This program, inaugurated in 1973, provides grants for the education and train- 
ing of persons with low incomes. 

Estimates of payments made under this program are based on 1974 disbursement 
data supplied by the Department of Health, Education, and Welfare. These data, 
tabulated by institution, street address, zip code, and State, were subsequently 
county coded and summed to county totals. The resulting distributions were used 
to allocate both the 1973 and 1974 State totals. 

Supplemental Security Income (SSI) Program, Federal Payments 

The new Federal program of Supplemental Security Income payments for the aged, 
blind, and disabled, administered by the Social Security Administration and 
financed from the general revenue of the Treasury, became operational January 1 , 
1974. The Federal payments, in combination with supplementary State payments 
(see State and local government transfers), have replaced payments under the 
Federal -State welfare system whereby matching Federal grants were made to States 
for benefit payments to persons eligible for Old Age Assistance, Aid to the 
Blind, and Aid to the Permanently and Totally Disabled. 

For personal income purposes, these Federal transfers are estimated in two 
parts: "SSI Basic Benefits" and "Hold-Harmless." The latter item refers to 
the section of the new law which requires those States already paying benefits 
above the new Federal minimum to maintain the level of payment in effect as of 
January 1972 for all eligible recipients. However, to minimize the financial 
burden of this mandatory supplementation, the law also provides that those 
States concerned need pay only an amount not to exceed their 1972 cash public 
assistance payments to adult welfare recipients and that the remaining necessary 
funds be provided by the Federal Government. 

Estimates of the "SSI Basic Benefits" and the "Hold-Harmless" payments were 

made by allocating the State totals among the counties by the distribution of 

the number of persons on the SSI rolls. This information was made available 
by the Social Security Administration. 

All Other Federal Transfer Payments, n.e.c. 

The remaining Federal transfers, with the exception of payments to nonprofit 
institutions, represent relatively minor dollar amounts. They include Panama 
Canal Construction Annuity Act payments, Area Redevelopment Act payments, inter- 
est subsidy payments on higher education loans, trade adjustment assistance, and 
benefit payments made under the Manpower Development Training Act, Job Corps, 
and the Education Exchange program. Because of the absence of specific geo- 
graphic data, estimates of these transfers were based on the county distribution 
of civilian population. 



34 

State and Local Government 

Benefits from Social Insurance Funds 

State and Local Government Retirement Benefits 

Representing nearly one-third of all State and local government transfers, 
these cash payments include not only monthly benefit payments but lump sum pay- 
ments and withdrawals as well. Since direct data are unavailable, county esti- 
mates were made using the REMD county estimates of State and local government 
payrolls as distributors of the State estimates. 

Efforts are being made to develop alternative allocators based on data more 
closely associated with retirement payments. However, additional testing is 
needed to determine the extent of their validity. 

Cash Sickness or Temporary Disability Payments 

These are weekly cash benefits from State-administered programs to insured 
workers unemployed because of nonoccupational illness or accident. This type 
of program is in effect in only four States — Rhode Island, California, New 
York, and New Jersey. County tabulations of actual dollar amounts are not 
available. County estimates of cash sickness payments were made by apportion- 
ing the State totals in accordance with the annual distribution of civilian 
population. 

Veterans' Benefits 



Veterans' Aid 



The major portion of the benefits included in this category are welfare payments 
made by State and local governments to indigent veterans. Since there are no 
direct data available, county estimates of veterans' aid are based on selected 
population data. State totals for 1969-74 were distributed to counties by the 
number of veterans of the post-Korean Conflict period, tabulations of which are 
available from the Veterans Administration. 

Veterans ' Bonuses 

Bonuses were paid to veterans of World War II, Korea and Vietnam. The amount 
of veterans' bonuses paid out in 1967-74 were based on the county distributions 
of veterans of the post-Korean Conflict period. 

Other State and Local Government Transfer Payments 
Direct Relief Payments 

This component had, in recent years, comprised slightly more than half of total 
State and local government transfers. County information on the amount of bene- 



35 

fit payments made, separately, for Aid to the Aged, Aid to Families with Depend- 
ent Children, Aid to the Blind, Aid to the Disabled, and General Assistance 
(maintenance payments only) were available annually from the State Departments 
of Welfare and/or the National Center for Social Statistics of the U.S. 
Department of Health, Education and Welfare. These data were summed and used 
to distribute a combined total of programs falling into the category of direct 
relief payments. The measurement of these components would have been improved 
by the construction of county distributions for each individual program. How- 
ever, this was not possible because of frequent changes in the structure and com- 
position of the various programs. 

The introduction of the new Supplemental Security Income Program in 1974 (see 
Federal Government transfer payments) has made this component obsolete. Bene- 
fit payments under the various programs are now grouped differently (see below). 

Supplemental Security Income (SSI) Program, State Supplementation 

This item includes the State mandatory supplementation payments made under the 
Hold-Harmless provision of the SSI law (discussed under Federal Government 
Transfer Payments, SSI Program) and State optional supplementation payments. 
This latter category derives from the provision of the SSI law that allows 
States to voluntarily make benefit payments above the Federal minimum require- 
ments through their own (State-administered) programs. 

The county estimates are based on county distributions of the number of persons 
enrolled in the SSI program, available from the Social Security Administration, 
which were used to allocate the State control totals. 

Aid to Families with Dependent Children (AFDC) 

Prior to 1974, this category of public assistance was estimated in combination 
with four other programs under the general heading "Direct Relief Payments." 
Three of these programs are now covered by SSI and are estimated for 1974 as 
such (see Federal Government Transfer Payments, SSI Program). AFDC still oper- 
ates under the Federal -State welfare system whereby matching Federal grants are 
provided to the States for money payments to persons in need of public 
assistance. 

Estimates for 1974 were based on benefit payments data made available by vari- 
ous State Departments of Welfare and by the National Center for Social Statis- 
tics of the U.S. Department of Health, Education and Welfare. 

General Assistance 

This is the fifth of the five programs that had formerly been estimated in 
aggregate as "Direct. Relief Payments." It is aid furnished by State or local 
governments to needy individuals or families who do not qualify for help under 
the Federally aided assistance programs. There is no Federal financial par- 
ticipation in this program. 



36 

With the introduction of SSI in January 1974, as described for AFDC, General 
Assistance is now estimated separately. The county estimates are based on 
direct data provided by the State Departments of Welfare. 

State and Local Medicare Benefits 

These transfers represent the premiums for the Supplementary Medical Insurance 
program of Medicare paid by State and local public assistance agencies for aged 
persons receiving welfare. County tabulations of persons enrolled in the Sup- 
plementary Medical Insurance program, obtained from the Social Security Admin- 
istration, were used to apportion State and local Medicare benefits to indi- 
vidual counties. 

Public Foster Home Care 

County estimates of payments for public foster home care for the period 1970-74 
were estimated by allocating the total amount expended by the State to individual 
counties on the basis of the number of children 18 years of age and under as 
reported in the 1970 Census of Population. The 1969 State expenditures for 
this item were allocated to the counties by the 1960 Census of Population dis- 
tribution of the number of children 14 years of age and under extrapolated 
forward to 1969 by the change in civilian population. 

State and Local Government Auto Depreciation Payments 

This item is similar to the auto depreciation component in Federal transfers. 
County estimates for all years are derived by allocating the State estimates of 
the amount of State and local government auto depreciation payments according 
to the county disbursements of State and local government payrolls. 

State and Local Government Payments to Nonprofit Institutions 

The estimates were made in three pieces--private foster home care payments, 
educational assistance payments, and "other payments." For private foster home 
care payments, State totals were distributed to counties by the same allocators 
as were used for estimating public foster home care payments. The amounts paid 
out by States for educational assistance and for "other payments" were allocated 
to counties by civilian population. 

Business Transfer Payments 

Business transfers account for about 4 percent of total transfer payments in the 
United States. The estimates, of varying quality at the national level, are of 
questionable validity at the State and county levels. There are no direct data 
available except at the national level and little relevant secondary information 
available to use as allocators. 



37 

Consumer Bad Debts 

County estimates were derived by apportioning the State totals in accordance 
with the county distributions of retail sales as published in the quinquennial 
Census of Business. 

Corporate Gifts to Nonprofit Institutions 

State totals were allocated to counties by civilian population. 

Other Business Transfers 

Included in this category are cash prizes, unrecovered thefts from business of 
cash and capital assets, and personal injury payments from business other than 
to employees. County estimates were made by distributing State totals in pro- 
portion to the civilian population. 

PERSONAL CONTRIBUTIONS FOR SOCIAL INSURANCE 

Contributions made by individuals under the various social insurance programs 
are excluded from personal income by handling them as explicit deductions. Pay- 
ments by both employees and nonfarm self-employed are included in the series. 

The employee portion covers contributions for Old Age, Survivors, Disability, 
and Health Insurance (OASDHI), State unemployment insurance, railroad retirement 
insurance, cash sickness insurance, and Federal, State, and local public employee 
retirement systems, as well as premium payments for Government Life Insurance. 
Contributions of the self-employed relate to OASDHI only. These were first made 
in 1952 under amendments extending coverage of the OASI system as of January 1, 
1951. In most cases, contributions of employees are withheld from payrolls; 
those made by self-employed individuals are paid annually with their Federal 
income tax returns. 

The personal contributions item in the State and county personal income series 
is the same as that which enters the national accounts except for an adjustment 
for contributions to retirement systems made by Federal employees stationed 
abroad. 

In addition to individuals' contributions for social insurance, State and 
county personal income excludes the contributions made on behalf of employees 
by their employers. Employee and employer contributions, though viewed as part 
of the total earnings of employees, are not actually received by them in the 
current period. In the national income accounts, they are recorded as receipts 
of social insurance funds in the government sector. The contributions made by 
self-employed persons are treated in the same way--even though paid by chem out 
of current income. 

Employee contributions to most programs of social insurance are collected in 
a manner similar to taxes on wages and salaries in that they are withheld at 



38 









^t- 


o 


IS 


O r- 


wiorooioin 


t^ — 


■* 




















r— 


o 




I s - ^f- 


.^ — MOC\ 


o «* «-j- 


o 


















i — i 


Ol 


» 


















cn 
















c. 


r— 


o 


*3 


ro en 


CO 












c 




00 












I— 




o 


















•k 1 — 


QJ 


C 
O 












<4- 






















ra 


x: 


■I — 




; 








o 






















0) 
Q. 


1— 


4-) 

3 












p 






















^CL 




X5 












C 






















ro 


• 


■i— 












QJ 




o 


O O CO 


OJ *d- (£ 


o o 


o en 


Lf) 




QJ 


S- 












u 




o 


in i — r-- 


Lf)C\JC 


cv 


O ^- oo 


O 


«N 


E 


4-4 












s- 


Oo 


, 


















.O 


a 


c 












QJ 


CO 


o 


ro o 


OJ 












rj.< 


u 


o 












L\. 


p— 


o 


















X3 
ro 




u 








j-> 


























+-> 


p— 


TO 








cu 




























ra 


C 








u 


* 




«3r 


OJ 




CTi -=3- 


OJ OJ CO c 


Lf5 


r>- 


S_ 


c 


o 






, 


c 


* 


* — ^ 


r-. 


C\J 


oo r- r-~ 


OJ LO o. 


CO LT 


r— 


(Tl 


i — - 


ro 


o 


'/I 






00 


TO 




o 


O0 


CO 


r-» 


LO OJ 


ro cp> c- 


ro tc 


(XJ 


<o 


LT) 


i — - 


CO 


5- 






p 


S- 


* 


r.3 


i — 


.?. 




h ff* #* 


*\ ^ 






r 


f\ 




. — 


S- 


QJ 






C 


3 


*t 


o 




i — 


r» 


oj oo 


<.£) OJ 


C\ 




LT) "^ 




E 


QJ 


Ql 






a j 


CO 


I s - 


»-. 




Ln 


^ 


«d- ro 


00 














Q. 








E 


c 


Co 


o 




r— 


















to 




a; 






>-, 


h— I 


r— ■ 


o 




n 




















Q.X 






TO 






o 




,_■ 


















O 


3 


p 






Q. 


1 — 


■o 


» — -* 






















4J 












.0 


c 


























OJ 


"+- 






QJ 


•[ — 


fO 


to 






















QJ 


^ 


o 






E 


'•'.J 




s- 






















U 


<c 








o 


o 


en 


TO 




LT) 


CO c 


C 


E 


QJ 






(_> 


to 


>.o 


1 — 




.-,.; 


r- 


Oo i— 


■vT ijd r-- 


ro r~- co 


r— 


r^- 


ro 




N 






c: 




ct, 


t — 




<dv 


o 


oj r- 


CTi I-- «* LT 


lo cr> to oo 


I — 


p 


•p- 






■i — 


5- 


t — 


o 


oo 


*\ 




r» n #\ 


«N #\ 






' 


c 




ro 


ro 


00 








o 




Q 


CO 


p— 


or. 


roocOi — 


r" 




OJ 


OJ 




.a 


x: 








p— 


'.- 


■r. 
.J) 




Oo 


Lf) 

I s - 


O 


OJ OJ 


1 — 












1 — - 


P 


QJ 

sz 






TO 
C 


CO 


ai 
























ro 


LO 


p 






o 


<r 


fj 
























U 


P 








oo 


o 


rd 










• « 














a 


• 


*i— 


C 


l+- 






5- 


1 — 


P 










o • 














• 


■ 


P 


QJ 


o 






CD 


p 












• • 














• 


• 


L0 


E 








CL 


3 












• • 














to 


• 


T"^ 


>i 


c 




• 




.a 


~i 










• • 














c 


01 


P 


ra 


o 




CD 


, — 


* r™ 


OJ 










* » 














o 


C 


rc 


a. 


■1 — 




c; 


i — 


5- 


+J 










» • 














■r • 


o 


4-1 




4-> 




-i — 


TO 


P 


•r— 










• • 














•*-> 


•1 — 


00 


i+- 


TO 




•a 




C 


C 










• • 














.3 


+J 




O 


u 




c: 


M- 


O 


13 










a • 














Q 


3 


CD 




■ p" 




3 


O 


C_) 


« 










* • 














•i — 

5- 


J3 


•i — 


oo 
QJ 






o 


E 


. — 


QJ 










• • 














+j 


S- 


> 


Q. 


•l — 






3 


<■!■) 


F 










• OJ 














C 


.!./' 


•k— 


>, 






L|- 


OO 


c 


O 










• u 














O 


C 


COP 


c 




o 




o 


o 










• c: 














u 


o 






TO 






QJ 


10 


c 










• ra 
















o 


If ■ 


OO 






QJ 


JZ 


5- 


I— 1 












S- 














+J 




o 


3 


CJ 




OO 


4-> 


qj 














3 














!~ 


to 




Q 


> 




3 




a. 


1 — ' 
TO 








QJ 
U 


















E 


CO 

CD 


QJ 

> 


"i — 






TO 
U 


E 
o 


'!-- 


C 








c 




•p— 














0.i 


C 


■i— 


"' 






'ij 


':- 


o 


C 

u. 








■•■: 




JC 








tr 






•i — 


O 


P 
U 


> 


o 




4-J 


Lj- 


CD 


i. 








3 




+J 








C 






+J 


•r— 


0J 


0J 






10 


10 


u 


QJ 








U" 




r™ 








c 


co 


« 


QJ 


to 


Q.JZ 


»* 




1 — 


c 


c: 


D. 








r: 




ro 








'I™ 


c 




5_ 




CO 


p 


S- 




TO 


o 


■ 










•r- 




QJ 








;- 


a 


• 




XI 


fc. 




QJ 




4-4 




-!-> 


i — 












jr 








3 


r 




CO 


LO 


01 


t. 


> 




O 


p 


:;. 


fO 




















j2 


+- 




0) 


ro 


Q. 


o 


QJ 




P 


3 


o 


+4 








TO 




-a 








■■I— 


3 




CD 


U 




4- 


J 






X3 


ex 


o 








»p- 




c 








5. 


-C 




>> 




QJ 




C 




i — 


■( — 


E 


l_ 








L 




ro 








+ - 


■r 




o -o 


JZ 


" 


X" 




QJ 


S- 


) 1 










O 










(/ 


c: 


5- 




r— 


c 


p 


C 






> 


4-4 




o 








1/ 




>, 






c 


c 


4-) 




a. 


rD 




o 


~s 




QJ 


c 


QJ 


J- 1 












-P 




co C 


tj 


C 




E 




c 


•1 — 


OJ 




i — 


o 


> 










S- 




•r~ 


• c: •!- 




C 




QJ 


QJ 


• 1 - 


+J 


{ — 






u 


-j — 


•1 








c 




r— 


• o +- 


•■(- 


u 






U 




U 


jQ 




S- 




P 


+J 






«-. 


14- 




•r- 


■ T- 3 


i. 






4-5 


C 


>i a 


TO 


• 


QJ 


1 — 


to 


c 












J2 


-l-J X2 


QJ 


OJ 




C 


re 


r— 


to 


4^ 


■a 


SZ 


TO 


i— 


QJ 








co 




ro 


CO 3 >i- 


E 


L 




QJ 


i- 


C 






OJ 


O) 


C 


QJ 


rz 








c 




CO 


CHS- 


a 


C 




E 


3 


o 


J3 


co 


p 


. — 


o 


£ v; 


o 






HH 


c 




•r— 


O •>— 4- 


S- 


a 




<.. 


CO 




O 


•i — 


TO 


X" 


00 




a. 






a 


1( _- 




T3 


•f- S- C 




s. 




S- 


c 


C71 


TO 


X 


E 




L 


i 


E 






h- 


4- 






+J +J c 


+J 


3 




?J 


■I— 


'~ 


QJ 


•:- s 


••- 


o 


uJ 




o 






---- 


3 




•* 


3 C O 


QJ 


tfl 


• 


> 




T- 






P 


p 


a. 


• 


o 








^:~ 




10 


JD o 


5. 


C 




a 


p 


r 


lf- 


C 


o^ 






U3 








u 


._, 




S- 


•■- o a 




r - 




CD 


e 


ra 


O 


•p- 


QJ 


■a 


4- 




>*> 






E 


i- 




o 


s- a 


a, 








a' 


a; 








-a 


o 


a 


..:. 






c 


P 




> 


4-> -a > 


, Q 


a 


• 


[..- 


E 


E 


Cn 


l/l 


Ol 


«:3 




r— ■ 








u 


c 




'r— 


c: a> c 


> 


,<+- 




f0 


>■> 




C 


4- 1 


c 




E 


.a 








c 


c 




> 


O >>,r- 


O 


*!— 


1 — 


u 


o 


O0 


-I — 


£Z 


•1 — 


4-4 


O 


re 








■k-- 


u 




i_ 


o o o 




1 


rr! 


o 


. — 


ra 


C 


13 


OJ 


a 




1 














3 


•— E 


a 




u 


1 — 


Q. 


J^ 


c 


O 


_Q 


<~ 


+J 










1 — 






to 


(U a<i 


E 


"+J 


o 




E 




•r- 


E 






u 










TO 


TO 






QJ E 


a 


C 


■ — 


-a 


QJ 


QJ 


cn 


TO 


10 


>> 


3 










C 


C 




rs 


>, OJ T3 




a 




c 


C 


r— 


QJ 




4-4 


ro 


o 










o 


C 




QJ 


O 1 ra 




E 


-a 


<n 


3 


XJ 


XJ 


OJ 


c 


E 


CD 










(,'": 


10 




cn 


r- H_ c 


to 


c: 


c 






rO 




p 


a' 




-a 










i_ 


<,_ 


i— ro 


Q.^- %. 


s. 


5- 


ro 


QJ 


0J 


4-> 


QJ 


J 


c 


r- - 












a 


a 


ro 


Eaii- 


a 


QJ 




-.-) 


4-> 




x: 


1 — 


o 


■1— 


5- 










D. 


c 


. s- -a 


UJ CO t- 


"C 


> 


QJ 


TO 


ro 


CO 


P 


O 


CL 


TO 


QJ 














QJ i— 


ro 


OJ 


a 


+J 


4-> 


+-> 


•i^ — 




L0 


E 


4-> 


4-> 










i — 




T3 O 


cc 


u_ 


o 


ro 


OO 


OO 


x: 


P 


..O 


o 


QJ 


o._ 










ro 


IS, 


<U 








+J 






1— 


TO 


TO 


u 


o 


<t 










4-4 


V 


U_ 








to 


























o 


■\L 
























* 


^^ 










I— 


_J 






















* 








* 


.^ — | 



39 

the source of disbursement under the laws of the Federal Government and of some 
State and local governments. However, the amounts of such taxes are counted 
as part of personal income--as though first received by the employee and then 
paid to government. This is in line with the overall definition of personal 
income as a bef ore-tax measure. Admittedly, the difference in treatment 
accorded withheld taxes and social insurance contributions is somewhat 
arbitrary. 

The general procedure used to estimate personal contributions to social insur- 
ance was to allocate State totals to the counties on the basis of payrolls or 
proprietors' income in the relevant category of income. There are several 
exceptions to this rule. Government Life Insurance (GLI) contributions were 
estimated by allocating State totals to counties on the basis of veteran 
population. State UI and cash sickness contributions were generally appor- 
tioned in accordance with selected population series from the decennial 
Censuses of Population. These contributions are made by employees in only 
a few States. 



40 

RESIDENCE ADJUSTMENT 

"Residents" Defined 

Personal income is defined as the current income received by residents of an 
area from all sources. The meaning of the term "residents" is therefore of 
primary importance since it delimits the economies of the various political 
subdivisions in terms of the summary personal income measure. As defined, 
residents include, in addition to individuals, a small percentage of nonprofit 
institutions and private trust funds which are treated as quasi-individuals, 
since they either function to serve individuals directly or are established 
in their behalf. As they are nonprofit in character, they are clearly dis- 
tinguishable from business enterprises. Nonprofit institutions include religi- 
ous organizations, social and athletic clubs, labor organizations, nonprofit 
schools and hospitals, charitable and welfare organizations, and other private 
nonprofit agencies furnishing services to individuals. 

The meaning of the term "resident individual" requires some explanation. Mili- 
tary personnel are considered residents of the State and county in which they 
are stationed, even though they may consider themselves permanent residents of 
another area. Similarly, Federal civilian employees are counted as residents 
of the State and county in which they live and work or from which they commute 
to work, regardless of the area they consider to be their permanent residence. 
By the same token, members of the Armed Forces and Federal civilian employees 
who are located away from their State or county of usual residence are not 
counted as residents of that State or county in the measurement of personal 
income. If they are located overseas, they are excluded altogether. Exclusion 
of income disbursed by the Federal Government to its personnel stationed out- 
side the continental limits of the United States constitutes the only difference 
between the State series and the national personal income measures carried regu- 
larly in BEA's national income and product accounts. 

While the conventions used in defining the residence of Federal Government per- 
sonnel are the most important, similar conceptual problems occur in defining 
several groups of residents engaged in activities in the private sector. BEA 
measures the labor income of seasonal migrant workers in the regions in which 
the work is performed. No attempt is made to adjust the data to reflect a con- 
cept of "permanent" or legal residence. 9/ Rather, the wages of seasonal migrant 
workers are combined with the wages of nonmi grants and are assigned to the 
States and counties where the workers live while the work is being performed. 
However, the wages of a worker on a short temporary assignment away from his 
home office or usual place of work are measured as if he had worked continu- 
ously at his usual place of employment. 

In short, individuals actually residing in a State or county, civilian and mili- 
tary personnel alike, are covered by the personal income measures: those living 
elsewhere, even though normally residents of the State or county, are not 



9/ An exception to this general approach is made for the wages paid to migrant 
workers in Alaska, most of whom have little connnection with the Alaskan 
economy. As: part of the residence adjustment procedures, these wages are 
assigned to the States and counties where the workers are assumed to have 
their "permanent" residences. 



41 

counted as residents. This generalization does not mean that the personal in- 
come estimates for a particular State or county include the income of tourists 
or others in temporary stay; these are not counted as residents. T he concept 
of residence is based essentially on physical rather than usual, permanent, or 
legal residence. It differs from the Census concept mainly in the treatment 
of seasonal and short-term workers. Whereas Census includes such workers at 
their usual place of residence (rather than where they were on April 1), BEA 
assigns the wages of these workers to the area where they resided while per- 
forming the work. 

In the case of quasi-individuals, the residence of a nonprofit institution is 
determined according to its geographic location. The residence of a private 
trust fund or fiduciary, however, is considered to be identical to that of its 
beneficiary. Since the income received by these entities is received on behalf 
of individuals, BEA is concerned with the location of the individual rather 
than the location of the fund. 



Residence Adjustment Procedure 

Li order to estimate the total and per capita income of residents in a local 
area, all the ccmponents of personal income must be put on a "where-received" 
(i.e., place-of-residence) basis. In the BEA local area income series, divi- 
dends, interest, rental income, and transfer payments are estimated on a where- 
received basis. The proprietors' income series is treated as being on a where- 
received basis although fragmentary information indicates that for many urban 
counties the estimates may represent a mixture of the place-of-work and place- 
of-residence classification. 

Wages and salaries, other labor income (OLI) and personal contributions were 
adjusted to convert them from a where-earned to a where-received basis. 
Excluded from adjustment were farm wages and OLI, private household wages 
and OLI, the allowances and allotments portion of military pay, the pay of 
military reservists (the latter a component of OLI), personal contributions 
for Government Life Insurance, State unemployment and cash sickness insurance, 
and contributions by the self-employed for Old Age, Survivors, Disability and 
Health Insurance, since these are already on a place-of-residence basis. [The 
term "income subject to adjustment" (ISA), referred to throughout this section, 
represents the sum of all the personal income components estimated on a place- 
of-work basis.] 

The residence adjustment was based primarily on commuting and income data 
obtained from the 1960 and 1970 Censuses of Population and on the 1972 county 
wages reported by the Internal Revenue Service. The; Census commuting data 
give the place of residence and place of work of employed persons at the State 
and county levels. Data on average incomes for selected groups of commuters 
were also available from the 1970 Census and were incorporated, for the first 
time, into the 1973 estimating cycle. The Census commuting and average income 
data were used in the procedures described below to yield the primary residence 
adjustment. Census county wage data for 1969 and IRS county wage data for 1972 
were used to identify the areas that needed further adjustment and to determine 
the size of the correction. 



42 

The method used in deriving the primary residence adjustment follows. First, 
an average "income subject to adjustment for residence" (ISA) was computed for 
each county for each year. These averages were calculated by dividing the ISA 
by the corresponding total employment, derived from the 1960 and 1970 Census 
commuting flow tabulations. The estimates of employment for the intercensal 
years were derived by a straight-line interpolation between the 1960 and 1970 
Census commuting data benchmarks. The 1970 Census commuting data were used for 
1971-74 (see Exhibit C). 

The average ISA's were weighted for each commuting flow using 1969 special 
tabulations of Census income data,J_0/ These Census data relate the average 
income of persons working in a given county, classified by county of residence, 
to the average income for all workers in that county of work. Thus, if the 
Census income data indicated that commuters working in County A and living in 
County B had an average income 1.5 times as great as all workers employed in 
County A, then the average ISA for County A was multiplied by 1.5 to obtain 
a weighted average income for that specific group of commuters. 

The weighted income average of each group of commuters (as described in the pre- 
ceding paragraph) was multiplied by the number of workers in that group (derived 
from the Census commuting flow tabulations) to yield a preliminary gross income 
flow: (The gross flows across State lines were adjusted to the controls esti- 
mated in the residence adjustment of State data.) All gross flows pertaining 
to a given county were then used to compute the county net residence adjustment; 
i.e.., the sum of the flows into the county minus the flows out of the county. 
(NOTE: In the foregoing procedure, the "employment" and "average income" data 
were derived for computational purposes only and were not intended to serve as 
measures of actual employment or average income of commuters.) 

For the years following 1970, under the foregoing procedure, all the dollar 
flows out of the county of work grow at the same rate as the ISA of that county. 
The resulting net flow of income between a given county and its neighbors varies 
from year to year because of the differences in the rates of growth among the 
counties. If a county's ISA is growing faster than that of its commuting 
neighbors, its not residence adjustmen t as a percent of its ISA becomes more 
negative. Conversely, if the growth is slower, the relative net adjustment 
becomes more positive (see Exhibit D). 

The estimates of total personal income, adjusted for residence as described 
above, were subjected to careful editing. The editing centered around the dis- 
tribution of residence-adjusted wages within BEA economic areas Y\J since there 
is little commuting across economic area boundaries. This editing provided an 
indication as to which counties within an economic area required further adjust- 
ment as well as the size of the adjustment. 

harp changes in employment and in commuting patterns can occur in the course 
of a few years, and such sudden changes can lead to significant errors in the 



10/ This was done for approximately 40 percent of all the commuting flows for 
which basic data are available from Census, including all flows of major 
importance. The remaining commuting flows were given a weight of one. 

IV See Appendix, "Classification of Economic Areas," for definition. 



43 

estimates of total personal income and per capita personal income. This prob- 
lem has, however, been handled with reasonable effectiveness by the use of 1969 
wage data from Census and 1972 wage data from the IRS in a comprehensive editing 
routine. Ad hoc adjustments were, for the most part, made only in those cases 
where correlative information supported IRS or Census data. (In counties domi- 
nated by military bases, the BEA estimates are consistently higher than those 
of either Census or IRS. They were left that way.) Where ad hoc commuting 
adjustments were justified, all available information about the changes in indus- 
trial or demographic structure of the county were utilized in determining the 
size of the additional adjustment. If the income estimates for a given county 
appeared to be too high, an adjustment would be made only if further analysis 
indicated that the estimate for an adjacent or nearby county was too low. 
Increasing the gross flow from the former to the latter then improved the 
estimates for both counties. 



PER CAPITA INCOME 

Per capita income is computed by dividing the residence-adjusted total personal 
income by population estimates. For 1971 forward, the population data are pro- 
vided by the Bureau of the Census for July 1 of each year. 12/ For earlier 
noncensal years, Census State population totals, available annually, were dis- 
tributed among the counties using county population data available from the 
individual States. 

BEA attempts, with varying degrees of success, to make its total personal income 
series geographically consistent over time. In some cases, this is difficult to 
do because of the annexation of territory of one jurisdiction by another, lead- 
ing to discontinuities in the population data. This leads to an understatement 
or overstatement of the per capita estimates for the years prior to annexation. 
In those instances where such errors are significant, the per capita income 
estimates are suppressed. 

Per capita income can vary widely from county to county. Those with either 
extremely high or low per capita incomes are generally the small counties. In 
many instances, an unusually high (or low) level of per capita income is tem- 
porary and results from unusual, conditions such as a bumper crop, a major con- 
struction project (i.e., a defense facility, a nuclear plant, or a dam), or 
a catastrophe (i.e., flood, tornado, or drought). In some cases, a high per 
capita income is illusory (for example, when a construction project brings in 
a large number of high-paid workers who live near the site and are included 
in the population count, but who send a substantial portion of their wages to 
their dependents living at their permanent homes in other counties). Conversely, 
counties with heavy institutional populations may show unusually low per capita 
incomes. 

In addition, because population is measured as of one date, whereas income is 
measured as a flow over the calendar year, a significant change in population 
during the year, particularly after July 1, can cause a distortion in the per 
capita figure. 



12/ However, these data reflect the college student population as of April 1 



44 

It should be noted that the considerable differences between BEA's estimates of 
per capita income and Census' estimates are due to differences in definition of 
income, collection mode, and method of computation. For example, Census com- 
putes 1969 per capita income by dividing 1969 total money income by 1970 total 
population, whereas BEA derives its 1969 per capita income by dividing 1969 
total personal income by 1969 total population. 



»U.S. GOVERNMENT PRINTING OFFICEU976 240-848/61 



Upr 




i